Filipin III: Configuration assignment and confirmation by synthetic correlation

Article abstract of DOI:10.1021/jo960218x

The stereochemical configuration of filipin III (1) was determined using the ¹³C acetonide analysis. The relative configurations for the nine stereogenic centers in the top half of filipin were initially identified using just three acetonide derivatives (2, 3, and 4) arising from a two-step protection sequence. The structure was confirmed by synthesis and direct correlation of degradation products 8 (C26-C28) and 10 (C1-C16). Filipin tetraacetonide 2 and triacetonide 4 each contain an anti acetonide in a highly unusual chair conformation. Molecular modeling successfully reproduced the preference for a chair conformation over the normally more stable twist-boat conformation.

Full text of DOI:10.1021/jo960218x

J . Org. Chem. 1996, 61, 4219-4231
4219
F ilip in III: Con figu r a tion Assign m en t a n d Con fir m a tion by
Syn th etic Cor r ela tion
Timothy I. Richardson and Scott D. Rychnovsky*
Department of Chemistry, University of California, Irvine, California 92717-2025, and University of
Minnesota, Minneapolis, Minnesota 55455-0431
Received February 1, 1996X
The stereochemical configuration of filipin III (1) was determined using the 13C acetonide analysis.
The relative configurations for the nine stereogenic centers in the top half of filipin were initially
identified using just three acetonide derivatives (2, 3, and 4) arising from a two-step protection
sequence. The structure was confirmed by synthesis and direct correlation of degradation products
8 (C26-C28) and 10 (C1-C16). Filipin tetraacetonide 2 and triacetonide 4 each contain an anti
acetonide in a highly unusual chair conformation. Molecular modeling successfully reproduced
the preference for a chair conformation over the normally more stable twist-boat conformation.
Although the flat structures of many polyene macrolide
antibiotics are known, the complete stereochemical as-
signments have been made in only a handful of cases.1
We have shown that syn and anti 1,3-diols can be easily
distinguished by 13C NMR analysis of the corresponding
acetonides: the syn isomers have methyl resonances at
ca. 19 and 30 ppm, whereas the anti isomers have methyl
resonances at ca. 25 ppm.2 We report herein the absolute
and relative configuration of filipin III (1), Figure 1.3 This
stereochemical assignment was established using, in
part, the 13C acetonide analysis and was confirmed by
synthesis of degradation fragments.
F igu r e 1. Structure and configuration of filipin III (1).
popular tool in cell membrane research as a probe for
cholesterol.
Filipin’s interaction in cell membranes is an area of
active research, and recently its use as a probe for
cholesterol has been questioned.6 Unlike amphotericin
B and nystatin A1 which form sterol-dependent ion
channels, filipin is thought to be a simple membrane
disrupter.7 However, there is no generally accepted
model that explains the action of filipin in cell mem-
branes. The most popular model suggests that filipin
binds 3-â-hydroxysterols in a 1:1 stoichiometry, forming
large aggregates that lie parallel to and in the center of
the lipid bilayer.8 It is these aggregates that form the
characteristic bulges observed in freeze-fractured or
negatively stained membranes. Other models suggest
that filipin and sterols aggregate at the membrane
surface9 or are localized in the upper layer of the
membrane and cause deformations due to an increase
in surface pressure.10 The characteristic bulges could be
caused by filipin-sterol aggregates or by filipin ag-
gregates alone. Staining of free cholesterol with filipin
is used clinically in the study and diagnosis of Type C
Niemann-Pick disease.11 The development of detailed
models for the interaction of cholesterol with filipin or
filipin aggregation has been hampered because, until
now, the configuration of filipin III was unknown.
Filipin belongs to a class of natural products known
as the polyene macrolide antibiotics. There are over 200
known representatives of this class, most of which are
produced by soil actinomycetes belonging to the genus
Streptomyces.4,5 Although they exhibit antifungal activ-
ity, most polyene macrolide antibiotics, with the excep-
tion of amphotericin B and nystatin A1, are too toxic for
therapeutic applications. However, filipin has become a
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
(1) Configurational assignment by X-ray: (a) Amphotericin B,
Mechlinski, H.; Schaffner, C. P.; Ganis, P.; Avitable, G. Tetrahedron
Lett. 1970, 3873. (b) Roxaticin, Maehr, H.; Yang, R.; Hong, L.-N.; Liu,
C.-M.; Hatada, M. H.; Todaro, L. J . J . Org. Chem. 1989, 54, 3816.
Configuration by degradation, NMR analysis, and partial synthesis:
(c) Mycoticin, Schreiber, S. L.; Goulet, M. T. Tetrahedron Lett. 1987,
28, 6001. (d) Schreiber, S. L.; Goulet, M. T.; Sammakia, T. Tetrahedron
Lett. 1987, 28, 6005. (e) Schreiber, S. L.; Goulet, M. T. J . Am. Chem.
Soc. 1987, 109, 8120. (f) Nystatin, Lancelin, J .-M.; Beau, J .-M.
Tetrahedron Lett. 1989, 30, 4521. (g) Prandi, J .; Beau, J .-M. Tetrahe-
dron Lett. 1989, 30, 4517. (h) Nicolaou, K. C.; Ahn, K. H. Tetrahedron
Lett. 1989, 30, 1217. Partial structure of lienomycin: (i) Pawlak, J .;
Nakanishi, K.; Iwashita, T.; Borowski, E. J . Org. Chem. 1987, 52, 2896.
(j) Roflamycoin, Rychnovsky, S. D.; Griesgraber, G.; Schlegel, R. J . Am.
Chem. Soc. 1994, 116, 2623. (k) Pentamycin, Oishi, T. Pure Appl. Chem.
1989, 61, 427. (l) Nakata, T.; Hata, N.; Suenaga, T.; Oishi T. In Abstract
of Papers, 30th Symposium on the Chemistry of Natural Products,
Fukuoka, Oct 1988, p 540. Configuration by NMR analysis: (m)
Pimaricin, Lancelin, J .-M.; Beau, J .-M. J . Am. Chem. Soc. 1990, 112,
4060. (n) Candidin, Pawlak, J .; Sowinski, P.; Borowski, E.; Gariboldi,
P. J . Antibiot. 1993, 46, 1598.
(2) (a) Rychnovsky, S. D.; Skalitzky, D. J . Tetrahedron Lett. 1990,
31, 945. (b) Evans, D. A.; Rieger, D. L.; Gage, J . R. Tetrahedron Lett.
1990, 31, 7099. (c) Rychnovsky, S. D.; Rogers, B.; Yang, G. J . Org.
Chem. 1993, 58, 3511.
(3) A prelimiary report of this work has appeared: Rychnovsky, S.
D.; Richardson, T. I. Angew. Chem., Int. Ed. Engl. 1995, 34, 1227-
1230.
(4) Omura, S.; Tanaka, H. In Macrolide Antibiotics: Chemistry,
Biology and Practice; Omura, S., Ed.; Academic Press: New York, 1984;
p 351.
(6) (a) Milhaud, J .; Mazerski, J .; Dufourc, E. J . Eur. Biophys. J .
1989, 17, 151. (b) Castanho, M. A. R. B.; Prieto, M. J . E. Eur. Biochem.
J . 1992, 207, 125.
(7) Bolard, J . Biochim. Biophys. Acta. 1986, 864, 257-304.
(8) De Kruijff, B.; Demel, R. A. Biochim. Biophys. Acta. 1974, 339,
57-70.
(9) Elias, P. M.; Friend, D. S.; Goerke, J . J . Histochem. Cytochem.
1979, 27, 1247-1260.
(10) Severs, N. J .; Robenek, H. Biochim. Biophys. Acta. 1983, 737,
373-408.
(11) Coxey, R. A.; Pentchev, P. G.; Campbell, G.; Blanchette-Mackie,
E. J . J . Lipid Res. 1993, 34, 1165-1176.
(5) Rychnovsky, S. D. Chem. Rev. 1995, 95, 2021-2040.
S0022-3263(96)00218-6 CCC: $12.00 © 1996 American Chemical Society

4220 J . Org. Chem., Vol. 61, No. 13, 1996
Richardson and Rychnovsky
Sch em e 1
Ta ble 1. Ra te of Exch a n ge betw een Aceton e a n d
2,2-Dim eth oxyp r op a n e Usin g 2.6 m M P P TS a t 23 °C
time (h)
acetone (2.1 ppm)a
2,2-DMP (1.21 ppm)a
2
5
1
1
16
7
20
48
72
1.2
2.9
3.8
1
1
1
a
The reaction between 2,2-DMP and d6-acetone was monitored
by 1H NMR integration, and the integral ratios are listed.
Filipin was isolated in 1955 from Streptomyces filipin-
ensis found in a sample of Philippine soil.12 Much later
it was shown to be a mixture of four components: filipins
I (4%), II (25%), III (53%), and IV (18%).13 The flat
structure of filipin III, the major component of the filipin
complex, was assigned in a series of degradation stud-
ies.14 Filipin I, which has been difficult to characterize,
is probably a mixture of several components each having
two hydroxyl groups fewer than filipin III.15 Filipin II
is 1′-deoxyfilipin III.16 Filipin IV is isomeric to filipin III
and is probably epimeric at C1′ or C3.17
13C-En r ich ed Aceton e-DMP . The 13C acetonide
method is useful in the stereochemical analysis of polyols,
but it is limited by the relative lack of sensitivity of 13C
NMR. When a natural product is precious, it is advanta-
geous to use 13C-enriched acetone to prepare the ac-
etonide derivatives. This gives a many-fold boost in
sensitivity and simplifies the assignment of the acetonide
methyl peaks in the 13C spectra. Normally a dehydrating
agent such as 2,2-dimethoxypropane (DMP) is required
to get good yields of acetonides from polyols. A mixture
of acetone and DMP was found to exchange upon treat-
ment with a mild acid like pyridinium p-toluenesulfonate
(PPTS). Acetone-d6 and DMP (2:1 volume ratio; 3.4:1 mol
thesis was carried out with 50% enriched [1,3-13C2]-
acetone as described above to enhance sensitivity and
simplify identification of the acetonide 13C methyl signals.
A highly purified sample of filipin III gave the same
mixture of products when subjected to the reaction
conditions, confirming that these derivatives arise from
the major component of the complex. The mixture of
triacetonides was acetylated and separated to give tri-
acetonide 3 in 64% yield and triacetonide 4 in 19% yield.
The structures of triacetonides 3 and 4, including the
positions of the acetates in each compound, were estab-
1
ratio) were mixed together with 2.6 mM PPTS. A H
NMR of the mixture was taken, and the ratio of inte-
grated peaks at 2.1 ppm (acetone) and 1.21 ppm (DMP)
was measured, Table 1. The results show that the
isotopic label was completely scrambled between acetone
and DMP after 72 h. This procedure was used to prepare
an enriched mixture of [1,3-13C]acetone and DMP for
derivatizing filipin III. The 13C-enriched solvent mixture
was recovered by evaporation and trapping and could be
reused.
Con figu r a tion of F ilip in III. Initial attempts to
prepare acetonide derivatives of filipin III using p-
toluenesulfonic acid led to extensive decomposition.
Treatment of filipin complex18 (Sigma) with a mild acid
catalyst, pyridinium p-toluenesulfonate (PPTS), in 2:1
acetone:DMP, followed by silica gel chromatography gave
tetraacetonide 2 in 6% yield and a mixture of triac-
etonides in 31% yield (Scheme 1). The acetonide syn-
1
lished using H NMR and COSY data.
1
Several resonances in the H NMR spectrum of filipin
and its acetonide derivatives 3 and 4 are easy to assign.
The proton at C2 is distinct, being the only proton to
resonate at ca. 2.5 ppm. In addition, the protons at C25-
C28 are each easy to assign because they rarely overlap
with other protons, and as a result the correlations
between them can be identified in a COSY spectrum.
1
In the H NMR spectrum of triacetonide 3 there are
three acetate methyls and six acetonide methyls. There
are four ester methine protons that resonate in the 5.3-
5.7 ppm range. Two of these can be assigned to the
protons at C26 and C27. Only one of these protons is
correlated to the proton at C2. This proton must be due
to an acetate at C1′ because if the acetate were at C3 it
would be impossible to account for three acetates and
three acetonides. The remaining proton is a doublet of
doublets. This proton must be due to an acetate at C15
because it is the only proton in the polyol segment that
could have such a simple splitting pattern. Therefore,
in triacetonide 3, the three acetates are at C1′, C15, and
C26 and the three acetonides are at C3-C5, C7-C9, and
C11-C13.
(12) (a) Whitfield, G. B.; Brock, T. D.; Ammann, A.; Gottlieb, D.;
Carter, H. E. J . Am. Chem. Soc. 1955, 77, 4799. (b) Ammann, A.;
Gottlieb, D.; Carter, H. E. Plant Dis. Rep. 1955, 39, 219.
(13) Bergy, M. E.; Eble, T. E. Biochemistry 1968, 7, 653.
(14) (a) Berkoz, B.; Djerassi, C. Proc. Chem. Soc. 1959, 316. (b)
Djerassi, C.; Ishikawa, M.; Budzikiewicz, H.; Shoolery, J . N.; J ohnson,
L. F. Tetrahedron Lett. 1961, 383. (c) Golding, B. T.; Rickards, R. W.;
Barber, M. Tetrahedron Lett. 1964, 2615. (d) Ceder, O.; Ryhage, R.
Acta Chem. Scand. 1964, 18, 558.
(15) Pandey, R. C.; Rinehart, K. L. J . Antibiot. 1970, 23, 414.
(16) Edwards, D. M. F. J . Antibiot. 1989, 42, 322.
(17) Pandey, R. C.; Narasimhachari, N.; Rinehart, K. L.;. Millington,
D. S. J . Am. Chem. Soc. 1972, 94, 4306.
1
In the H NMR spectrum of triacetonide 4 there are
three acetate methyls and six acetonide methyls. There
are four ester methine protons that resonate in the 5.3-
5.7 ppm range. Two of these can be assigned to the
protons at C26 and C27. Only one of these protons is
correlated to the proton at C2. As in triacetonide 3, this
(18) Filipin complex was obtained from Sigma (F-9765) as a crude
mixture containing ∼50% filipin III or from Upjohn (lot no. 2923-DEV-
39) containing about ∼8% filipin III. It can be purified by reversed-
phase preparative HPLC.

Filipin III
J . Org. Chem., Vol. 61, No. 13, 1996 4221
F igu r e 2. Chemical shift (13C) of 2,2,4,4,6-pentamethyl-1,3-
dioxane which adopts a chair conformation.
proton must be due to an acetate at C1′ because if the
acetate were at C3 it would be impossible to account for
three acetates and three acetonides. The remaining ester
methine proton has a complex (dddd) splitting pattern.
This proton cannot be due to an acetate at C15 or C3. It
can only be assigned to C7 or C11. Fortunately, the
F igu r e 3. Chair and twist-boat conformation of methyl and
vinyl acetonides at RHF/6-31G*//RHF/6-31G*.
Sch em e 2
1
protons in the H NMR spectrum of this compound are
well separated, and using COSY data we were able to
assign this acetate to C7. Thus the triacetonide 4 has
the three acetates at C1′, C7, and C26 and the three
acetonides at C3-C5, C9-C11, and C13-C15.
The 13C NMR spectrum of tetraacetonide 2 shows
methyl signals at 31.0, 30.7, 30.5, 30.1, 24.8, 20.0, and
19.5 (×2) ppm. These signals correspond to three syn
acetonides and one acetonide with methyl peaks at ca.
30 and 24.8 ppm. The 13C NMR spectrum of triacetonide
3 shows methyl resonances at 30.8, 30.7, 30.0, 19.9, 19.8,
and 19.4 ppm that correspond to three syn acetonides.
The 13C NMR spectrum of triacetonide 4 shows methyl
resonances at 31.1, 30.5, 30.2, 24.7, 19.0, and 19.0 ppm.
These signals correspond to two syn acetonides and one
acetonide with methyl peaks at ca. 30 and 24.7 ppm. The
overlapping acetonide rings in compounds 2, 3, and 4
cover all of the relative 1,3-diol relationships in the C1′
to C15 segment of filipin III.
A logical analysis of the three acetonide derivatives,
2-4 in Scheme 1, can be used to assign most of the
stereochemical relationships in filipin III. Compound 3
showed only syn acetonides, so C3-C5, C7-C9, and
C11-C13 were assigned syn. Compound 4 showed one
anti acetonide, but C3-C5 was previously assigned as
syn, so the anti ring must be either C9-C11 or C13-
C15. Compound 2 showed only one anti ring and also
incorporates both C9-C11 and C13-C15 acetonides. It
has been established that one of these must be anti. With
the anti ring accounted for, both C5-C7 and C1′-C3
must be syn. Combined with a straightforward analysis
of the C2 configuration (see below), the 13C acetonide
analysis has reduced the number of possible diastereo-
mers for the C1-C15 section of filipin III from 256 to 2.
Both compounds 2 and 4 show a very unusual set of
methyl peaks for one of the acetonide rings. Over 200
acetonide rings were analyzed in a previous paper, and
none of them showed a pair of methyl peaks at the
unusual positions of ca. 25 and 30 ppm.2c One compound
that does show similar 13C chemical shifts for the
acetonide methyl groups is 2,2,4,4,6-pentamethyl-1,3-
dioxane, Figure 2. It is known to adopt a chair confor-
mation because the quaternary center at C4 disfavors the
twist-boat conformation.19 The C2 axial and equatorial
13C methyl groups appear at 24.93 and 32.04, respec-
tively.19 These are very close to the observed values of
ca. 30 and 24.8 or ca. 30 and 24.7 ppm in compounds
2 and 4, respectively. Apparently the macrocyclic ring
biases the anti 1,3-diol acetonide to adopt a chair
conformation rather than the more stable twist-boat
conformation. The chair conformation, although still
energetically unfavorable, is more accessible for an
alkene-substituted acetonide than for an alkyl-substi-
tuted acetonide, Figure 3.20 Thus the most likely position
of the anti acetonide is at C13-C15, which has an alkene
at C16-C17 that can occupy an axial position on the
acetonide ring.
To confirm the tentative conclusion that the anti diol
in filipin III was at C13-C15, compounds 2 and 4 were
degraded as shown in Scheme 2. Acetylation, followed
by hydrogenation of 2, gave the saturated macrocycle 5,
which showed methyl signals at 30.75, 30.69, 30.14,
25.69, 24.78, 19.95 (×2), and 19.45 ppm. The methyl
peaks at 25.69 and 24.78 ppm fall in the expected range
for a saturated anti acetonide. Triacetonide 4 was
ozonized and reduced with triphenylphosphine to give
triacetonide 6, which showed methyl signals at 30.50,
30.30, 25.94, 24.32, 19.57, and 19.29 ppm. The methyl
peaks at 25.94 and 24.32 ppm clearly represent an anti
1,3-diol acetonide. More significantly, the 1.62 ppm
difference between the two peaks is larger than one
would expect for a saturated anti 1,3-diol acetonide but
is in line with an acyl-substituted anti acetonide.2c Thus
the anti acetonide ring in both 2 and 4 is located at C13-
C15, and the relative configurations for all the alcohol
stereogenic centers in the C1′-C15 section are as shown
in Figure 4.
The acetonide at C1′-C3 adopts a chair conformation
in tetraacetonide 2. The C2 proton in the 1H NMR
(19) Pihlaja, K.; Kivimaki, M.; Myllyniemi, A.-M.; Nurmi, T. J . Org.
Chem. 1982, 47, 4688.
(20) Rychnovsky, S. D.; Yang. G.; Powers, J . P. J . Org. Chem. 1993,
58, 5251-5255.

4222 J . Org. Chem., Vol. 61, No. 13, 1996
Richardson and Rychnovsky
F igu r e 5. ∆δH ) (δS - δR) for the Mosher esters of 7 and 8
(in Hz) at 500 MHz.
F igu r e 4. Overview of the configurational assignment of
filipin III.
Sch em e 3
F igu r e 6. Retro-synthetic analysis of degradation fragment
10.
idene-1,2,3-butanetriol.21 The advanced Moshers analy-
sis is reported in Figure 5.22 Reduction of 7 gave a
mixture of acetonides that was treated with acetone and
acid to give pentaacetonide 9 in 76% yield. The 13C NMR
spectrum of 9 showed four syn 1,3-diol acetonides and a
mixture of syn and anti 1,2-diol acetonides, which lends
independent support for the relative configurations pro-
posed in Figure 4.
The last question to be resolved is the absolute stere-
ochemistry of the C1′-C15 fragment. An advanced
Mosher ester analysis22 of the C1′ alcohol of pentaac-
etonide 7 gives the ∆δH () δS - δR) values listed in Figure
5. These values indicate an (R) configuration at C1′, and
complete the stereochemical assignment of filipin III.
Thus the configurations of the stereogenic centers in
filipin III are 1′R, 2R, 3S, 5S, 7S, 9R, 11R, 13R, 15S,
26S, and 27R.
Syn th esis a n d Cor r ela tion of Degr a d a tion F r a g-
m en t 10. Pentacetonide 10, one of the degradation
fragments from the stereochemical assignment of filipin
III, was identified as an excellent synthetic target
because it contains nine of the stereogenic centers present
in the natural product. Pentaacetonide 9 is a mixture
of epimers at the C16 center, and the C15/C16 anti
isomer 10 was selected as the synthetic target. Correla-
tion of the degradation product 10 with a sample
prepared by total synthesis would confirm the relative
and absolute assignment of filipin III’s C1-C15 polyol
segment.
spectrum of tetraacetonide 2 is a triplet with a diaxial
coupling constant of 10.1 Hz. Therefore, the C2 ester is
anti with respect to the hydroxyls at C1′ and C3.
The configuration of filipin III at the C26 and C27
stereogenic centers was determined by another degrada-
tion sequence (Scheme 3). Per-TES-filipin prepared from
filipin complex (Upjohn or Sigma) was ozonized and then
reduced with NaBH4. The triethylsilyl (TES) protecting
groups were exchanged for acetonides to give 15-22% of
pentaacetonide 7 as a mixture of diastereomers at C16.
The position of the lone hydroxyl was identified by COSY
analysis of acetylated 7. LAH reduction of 7 gave a
quantitative yield of anti-1,2-di-O-isopropylidene-1,2,3-
butanetriol (8) that was identified by direct comparison
(1H NMR, GC) with an authentic sample of the enanti-
omer prepared according to eq 1.21 The absolute config-
A retrosynthetic analysis of pentaacetonide 10 is shown
in Figure 6. The proposed synthetic strategy relies on
the alkylation and stereoselective reductive decyanation
of cyanohydrin acetonides, a methodology designed to
facilitate the convergent synthesis of alternating polyol
chains.23 Using this strategy, pentaacetonide 10 was
uration of 8 was identified as (2S,3R) by comparison of
the (R)- and (S)-Mosher’s esters of 8 with those prepared
from an authentic sample of (2R,3S)-1,2-di-O-isopropyl-
(22) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J . Am.
Chem. Soc. 1991, 113, 4092-4096.
(23) Rychnovsky, S. D.; Griesgraber, G. J . Org. Chem. 1992, 57,
1559-1563.
(21) Mulzer, J .; Angermann, A.; Mu¨nch, W.; Schlichtho¨rl, G.; Hentz-
schel, A. Liebigs Ann. Chem. 1987, 7.

Filipin III
J . Org. Chem., Vol. 61, No. 13, 1996 4223
Sch em e 4
Sch em e 5
Sch em e 6
conveniently divided into two fragments, left-hand frag-
ment 11 and right-hand fragment 12.
Syn th esis of Left-Ha n d F r a gm en t 11. A model
study was employed to test the possibility of installing
the 1,2-diol late in the synthesis, Scheme 4. We envi-
sioned that the nascent diol would be introduced as an
alkyne and carried through the construction of the
alternating polyol. The alkyne would be reduced to an
(E)-alkene concurrently with the final reductive decya-
nation. Sharpless asymmetric dihydroxylation (AD) of
the (E)-alkene would then be used to install the required
1,2-diol at C15-C16 with reagent-controlled stereoselec-
tivity.
in Scheme 5.25 Unfortunately crotyl chloride (20) is only
available as a 5:1 mixture of E and Z isomers. Sharpless
asymmetric dihydroxylation of 20 gave an 86% yield of
chloro diol 21 which was converted to acetonide 22 under
standard conditions. GC analysis of 22 showed a 6:1
mixture of trans and cis isomers. The trans isomer
showed 81% ee by GC analysis using a chiral column.
The chlorine of 22 was then exchanged for iodine by
treatment with KI and 18-crown-6 in refluxing xylenes
for 4 days. The trans isomer was selectively converted
into the iodide: the mixture was composed of 65% of the
iodide that was a 97:3 mixture favoring the trans isomer,
while the recovered chloride made up 35% of the mixture
and was as a 2:1 mixture favoring the trans isomer. The
1,2-diol synthon 23 was available in three steps as a 97:3
mixture of diastereomers with 81% ee.
The synthesis of the left-hand fragment was completed
as shown in Scheme 6. Chloro cyanohydrin 13 (2 equiv)
was deprotonated with lithium diethylamide and alky-
lated with iodide 23. Diacetonide 24 was isolated in 76%
yield (based on the iodide) as an 80:20 mixture of
diastereomers26 which was used without further purifica-
tion. Chloride 24 was converted to iodide 25, and
alkylation of (2 equiv) chloro cyanohydrin anion 13 with
25 gave triacetonide 26 in 94% yield based on 25.
Triacetonide 26 was recrystallized to give a white crys-
talline solid, mp ) 126-127 °C, in 70% yield with a
diastereomeric purity of >97%.27 The structure of 26 was
Chloro cyanohydrin acetonide 13 was prepared from
ethyl 4-chloroacetoacetate (54% overall yield, 91% ee).23
It was alkylated with 1-chloro-2-butyne in 58% yield to
give nitrile 14. The chlorine of 14 was then exchanged
for iodine using KI and 18-crown-6 in refluxing xylenes
for 3 days.23 This gave a near quantitative recovery of
material, but 1H NMR showed only 79% conversion to
the iodide. The iodide was then reduced with tributyltin
hydride and AIBN in refluxing toluene. This gave a
quantitative yield of nitrile 16 contaminated with 20%
of the chlorine-containing nitrile 14. The nitrile and the
alkyne of 16 were reduced with Li in liquid ammonia to
give alkene acetonide 17 as a single isomer in 73% yield.
Sharpless asymmetric dihydroxylation of 17 gave a 58%
1
yield of diol 18. However, the 13C and H NMR spectra
of 18 showed it to be a mixture of diastereomers. The
diol was acetylated to give a 99% yield of diacetate 19.
GC analysis of this product revealed a 68:32 mixture of
diastereomers that was not resolvable by silica gel
chromatography. The poor selectivity in the dihydroxy-
lation is not without precedent; as Sharpless noted in a
recent review, the selectivity is often compromised if the
“large” substituent of the target olefin is oxygenated.24
The model study demonstrated that the asymmetric
dihydroxylation would not be selective in an advanced
intermediate. The diol was instead introduced at the
beginning of the synthesis using the previously reported
asymmetric dihydroxylation of crotyl chloride as shown
(25) Vanhessche, K. P. M.; Wang, Z.-M.; Sharpless, K. B. Tetrahe-
dron Lett. 1994, 35, 3469.
(26) The ratio was determined by 1H NMR integration of the major
C16 methyl doublet at 1.29 ppm versus the minor C16 methyl doublet
at 1.26 ppm.
(24) Kolb, H. C.; VanKieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483-2547.
(27) Diastereomeric purity determined by 1H NMR analysis.

4224 J . Org. Chem., Vol. 61, No. 13, 1996
Richardson and Rychnovsky
Sch em e 8
F igu r e 7. Crystal structure of 26 showing the relative and
absolute configuration.
Sch em e 7
31 with OsO4/NMO followed by NaIO4 gave aldehyde 32
in 99% yield. Addition of dIpc2BAllyl to 32 provided
homoallylic alcohol 33 which was isolated as a single
diastereomer in 86% yield. The minor diastereomer was
separated and isolated in 5% yield. Synthesis of the
right-hand fragment was completed by protection of
homoallylic alcohol 33 as the TMS ether followed by
oxidative cleavage of the alkene to the aldehyde, cyano-
hydrin formation, and ketalization to provide cyanohy-
drin acetonide 12 in 74% yield from alcohol 33.
F in a l Cou p lin g a n d Com p letion of th e Syn th etic
Cor r ela tion . The final steps in the synthesis of pen-
taacetonide 10 are shown in Scheme 8. Cyanohydrin
acetonide 12 (3 equiv) was deprotonated and alkylated
with iodide 27 to give pentaacetonide trinitrile 34 in 66%
yield based on the iodide. Stereoselective reductive
decyanation of 34 provided pentaacetonide 10 in 79%
yield. The 1H NMR spectrum of synthetic pentaacetonide
10 was identical to that of one of the isomers of the
degradation product 9. The rotation of synthetic 10,
[R]26 ) -19.2° (c ) 1.05, CHCl3), compares favorably
D
with the rotation of the degradation product 9, [R]26
)
D
-20.5° (c ) 1.43, CHCl3). Although the degradation
product is a 1:1 mixture of 10 and the C16 epimer, the
very similar rotations clearly support the proposal that
the synthetic 10 and degradation product 9 have the
same absolute configuration.
Molecu la r Mod elin g of th e Ch a ir ver su s Tw ist-
Boa t Con for m a tion of th e C13-C15 Aceton id e. The
anti acetonide rings in both tetraacetonide 2 and tri-
acetonide 4 adopt chair conformations on the basis of the
NMR analysis described above. Are these conformational
preferences reproduced by molecular modeling? Can
molecular modeling predict which acetonides will adopt
such unusual conformations?
Previous studies with MM2* have shown that it does
not reproduce chair versus twist-boat preferences well.20
Since that time the MM2* force field in MacroModel has
been reoptimized to give better acetal structures and
energies.31 The chair versus twist-boat preferences
calculated with the new MM2* force field in MacroModel
5.0 are shown in Table 2, and are compared with MM2*
using Macromodel 3.5 and experimental results.20 These
results show that the new force field is less biased toward
confirmed by X-ray crystallography and is illustrated in
Figure 7.28 Chloride 26 was converted into iodide 27 to
complete the synthesis of the left-hand fragment.
Syn th esis of Righ t-Ha n d F r a gm en t 12. Synthesis
of the right-hand fragment is illustrated in Scheme 7.
The asymmetry of right-hand fragment 12 was generated
using a Sharpless asymmetric epoxidation of (Z)-2-octen-
1-ol (29). Oxirane 30 was isolated in 96% yield and 88%
ee.29 Oxirane 30 was opened with vinylmagnesium
bromide in the presence of CuI to give the expected
alkenediol.30 When this reaction was carried out on a
small scale, a 77% yield of a 1:4 mixture of 1,2- and 1,3-
diols was obtained. The diols were separated by MPLC
on silica gel and characterized individually. When this
reaction was carried out on a larger scale, the 1,2-diol
was cleaved with NaIO4, and the 1,3-diol was easily
purified and isolated in 50% yield. The 1,3-diol was
converted to acetonide 31 in 98% yield, and treatment of
(28) We thank Victor G. Young for determining the crystal structure
of 26. The author has deposited atomic coordinates for this structure
with the Cambridge Crystallographic Data Centre. The coordinates
can be obtained, on request, from the Director, Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge, CB2 lEZ, U.K.
(29) Enantiomeric purity by analysis using a chiral GC column. See
Experimental Section for details.
(31) Mohamadi, F.; Richards, N. G. J .; Guida, W. C., Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J .
Comput. Chem. 1990, 11, 440-467.
(30) Tius, M. A.; Fauq, A. H. J . Org. Chem. 1983, 48, 4131-4132.

Filipin III
J . Org. Chem., Vol. 61, No. 13, 1996 4225
Ta ble 2. Ca lcu la ted ∆H for th e Ch a ir to Tw ist-Boa t
Equ ilibr iu m of tr a n s-1,3-Dioxa n es (k ca l/m ol). Th e MM2*
F or ce F ield in Ma cr om od el 3.5 a n d 5.0 Ar e Com p a r ed
chair conformations than the old force field, but it still
predicts that the aldehyde, ester, and vinyl groups will
slightly favor the chair conformation. The experimental
results show that each one favors a twist-boat conforma-
tion. The predicted equilibrium for the methyl case
appears to be worse than before, but this is deceptive as
the methyl equilibrium has a large experimental uncer-
tainty. The calculated methyl equilibrium is in good
agreement with the 6-31G* value of -2.79 kcal/mol.20 The
vinyl group is the most relevant example for tetraac-
etonide 2, where MM2* shows about a 1 kcal/mol bias in
favor of the chair conformation. The new MM2* force
field is much better than the old in its handling of acetals
and in these simple cases reproduces the experimental
values to within about 1 kcal/mol.
F igu r e 8. Minima of filipin tetraacetonide with C13-C15
chair (top) and C13-C15 twist-boat (bottom) conformations.
Macromodel energies (kcal/mol) using MM2* with chloroform
solvation are -102.24 for the chair and -98.64 for the twist-
boat. Hydrogens have been omitted for clarity.
Filipin tetraacetonide 2 has many internal degrees of
freedom, and therefore an exhaustive conformational
search would be prohibitive. To model 2, each acetonide
ring was minimized separately and was not varied in the
Monte Carlo conformational search. The polyene is most
stable in the all-trans conformation, and internal torsions
in the polyene region were ignored, as were torsions in
the ester and the n-pentane side chain. The C13-C15
acetonide was built in as a chair or a 2,5 twist boat and
not varied. Separate Monte Carlo conformational searches
for both the chair and twist-boat starting geometries were
setup in Macromodel 5.0 using MM2* with 11 variable
torsional angles and were run for 5000 steps. The
minima were resubmitted to another 5000-step confor-
mational search that included some variable torsions in
the side chain. All of the combined structures were
reoptimized using the chloroform solvent set. The lowest
energy minima for both the chair and twist-boat geom-
etries were each found multiple times and are shown in
Figure 8. The chair conformation shows 9 minima within
1.0 kcal/mol, while the twist-boat conformation shows 14
minima within 1.0 kcal/mol, showing that both minima
are members of families of related structures. The lowest
energy chair conformation is favored over the lowest
energy twist-boat conformation by 3.6 kcal/mol. The
MM2* bias for chair conformations is about 1 kcal/mol,
but the other 2.6 kcal/mol can be attributed to a real
preference for the chair conformation in the filipin ring
system.
compound such as 6 the twist-boat is favored, but in a
cyclic compound such as 2 the overall strain is reduced
by adopting a chair conformation. Anti acetonides that
adopt a chair conformation are very unusual, but the
current example shows that they do occur. Molecular
modeling with the improved MM2* force field should be
a valuable tool to predict where other anti acetonide
chairs might be found.
Con clu sion
The relative and absolute configuration of filipin III
was determined using the 13C acetonide method in
conjunction with the Mosher method. The assignment
was confirmed by synthesis and spectroscopic correlation
of degradation fragments 8 (C26-C28) and 10 (C1-C16).
It must be emphasized that the relative configurations
for the nine stereogenic centers in the top half of filipin
were initially identified using just three acetonide de-
rivatives, 2, 3, and 4, arising from a two-step protection
sequence. Subsequent experiments confirmed the analy-
sis. The stereochemical assignment of filipin III will
facilitate rational modeling of filipin-cholesterol interac-
tions.
Tetraacetonide 2 contains an unusual anti acetonide
that adopts a chair conformation. Our published litera-
ture survey found no examples of anti acetonides with
two sp3 substituents that adopted a chair conformation.2c
Molecular modeling correctly predicts that the chair
conformation would be favored over the twist-boat con-
formation.
The C13-C15 anti acetonide ring in tetraacetonide 2
acts as a hinge. Figure 8 shows that the chair conforma-
tion is well placed to “turn the corner” on the macrocyclic
ring, whereas the twist-boat conformation points the two
ends of the chain in opposite directions and does nothing
to facilitate closure of the macrocyclic ring. In an acyclic

4226 J . Org. Chem., Vol. 61, No. 13, 1996
Richardson and Rychnovsky
Exp er im en ta l Section 32
EtOAc. The more polar material was taken up in the
recovered solvent. PPTS (3 mg) was added and the solution
allowed to stir overnight. The reaction was quenched with 5
µL of Et3N and the solvent evaporated under reduced pressure
to give a yellow solid. More of the triacetonide mixture (2 mg)
was isolated by silica gel chromatography (2 × 6 cm) eluting
with 30-50% EtOAc/hexanes (5% increments, 50 mL each
time). Total yield for the triacetonide mixture was 13.0 mg
(16.8 µmol, 31%).
Degr a d a tion Stu d ies on F ilip in III. F ilip in III (1).
Filipin complex (Sigma no. F 9765) (11 mg) was taken up in a
minimum amount of DMF and diluted with 65% MeOH/35%
H2O to make a 20 mg/mL solution. This solution was injected
(20-40 µL each time) onto a reversed-phase HPLC column and
eluted with 65% MeOH/35% H2O at 2.25 mL/min. The major
peak eluting at 18 min was collected to give 5.4 mg (8.2 µmol,
49%) of a slightly yellow solid: 1H NMR (500 MHz, CD3OD) δ
6.39 (dd, J ) 13.9, 11.2, 1 H), 6.33 (dd, J ) 8.9, 14.8 Hz, 1 H),
6.33-6.18 (m, 5 H), 5.94 (d, J ) 11.2 Hz, 1 H), 5.89 (dd, J )
5.11, 14.8 Hz, 1 H), 4.72 (q, J ) 6.3 Hz, 1 H, obscured by
MeOH), 4.08 (ddd, J ) 3.8, 7.4, 7.1 Hz, 1 H), 4.03 (dd, J ) 4.5,
10.8 Hz, 1 H), 3.97 (dd, J ) 6.3, 5.1 Hz, 1 H), 3.94-3.85 (m, 4
H), 3.75 (ddd, J ) 2.1, 8.4, 8.8 Hz, 1 H), 3.12 (ddd, J ) 3.4,
10.6, 10.5 Hz, 1 H), 2.45 (dd, J ) 7.4, 8.8 Hz, 1 H), 1.79 (ddd,
J ) 3.4, 9.4, 10.8 Hz, 1 H), 1.67 (s, 3 H), 1.66-1.58 (m, 3 H),
1.32-1.44 (m, 8 H), 1.16-1.30 (m, 8 H), 1.19 (d, J ) 6.3, 3 H),
0.816 (t, J ) 7.1 Hz, 3 H); HRMS (FAB) calcd for C35H59O11
655.4058, found 655.4020 (M + H).
Tetraacetonide 2: 1H NMR (500 MHz, C6D6) δ 6.78-6.67
(m, 2 H), 6.57-6.51 (m, 2 H), 6.48-6.42 (m, 2 H), 6.37-6.25
(m, 2 H), 5.87 (dd, J ) 5.0, 15.3 Hz, 1 H), 4.89 (dq, J ) 6.0, 6.4
Hz, 1 H), 4.48 (ddd, J ) 2.6, 10.0, 10.0 Hz, 1 H), 4.42-4.36
(m, 2 H), 4.28-4.22 (m, 2 H), 4.06-4.01 (m, 2 H), 3.90-3.84
(m, 1 H), 3.75 (ddd, J ) 4.1, 6.0, 7.4 Hz, 1 H), 2.49 (t, J ) 10.1
Hz, 1 H), 2.26 (d, J ) 13.8 Hz, 1 H), 2.12-1.20 (m, 19 H), 1.85
(s, 3 H), 1.543 (s, 3 H), 1.538 (s, 3 H), 1.50 (s, 3 H), 1.49 (s, 3
H), 1.46 (s, 3 H), 1.43 (s, 3 H), 1.34 (s, 3 H), 1.212 (s, 3 H),
1.206 (d, J ) 6.4 Hz, 3 H), 0.869 (t, J ) 6.9 Hz, 3 H); 13C NMR
(125 MHz, C6D6, isotopically enriched acetonide methyls) δ
31.0, 30.5, 30.5, 30.1, 24.8, 20.0, 19.5 (2); HRMS (FAB) calcd
for C47H74O11 814.5233, found 814.5239 (M).
1′,3,5,7,9,11,13,15,16-O-Non a k is(t r iet h ylsilyl)filip in .
Crude filipin complex (Upjohn U-5956 no. 2923-DEV-39) (2.0
g) and imidazole (5.2 g) were placed in a flask and flushed
with argon. TESCl (9.3 mL) was dissolved in 50 mL of DMF,
degassed with argon, and then transferred via cannula into
the reaction flask. The reaction was stirred under argon in
the dark for 9 h. The reaction was quenched with 50 mL of
saturated NH4Cl, and then the solution was extracted (3 ×
50 mL) with Et2O. The organic solution was washed with
saturated NaHCO3 (2×), water (2×), and brine, dried (MgSO4),
and then concentrated under reduced pressure to give a bright
orange oil. The least polar material which fluoresced under
long UV light was purified by silica gel flash chromatography
(5 × 11 cm) eluting with 2% Et2O/pentane to yield a yellow
oil. The major product in this mixture was further purified
on a silica gel MPLC column (4 × 30 cm) eluting with 2% Et2O/
pentane to give 470 mg (280 µmol, 9%) of per-TES-filipin as a
yellow oil.
1′,3:5,7:9,11:13,15-Tet r a k is-O-(1-m et h ylet h ylid en e)fil-
ip in III (2). Acetone (0.2 mL), [1,3-13C2]acetone (0.2 mL), and
DMP (0.2 mL) were combined in 2.0 mL of THF and degassed
with argon. PPTS (4.0 mg) was added and the solution allowed
to stir 64 h. Filipin Complex (Sigma) (35 mg, 54 µmol) was
added. The reaction was allowed to stir in the dark under
argon for 6 h. The reaction was quenched with 4 mg of solid
NaHCO3. The solvent was evaporated under vacuum and
recovered in a liquid nitrogen cooled trap. The solids were
taken up in 30 mL of EtOAc. The organic solution was washed
with saturated NaHCO3, water, and brine, dried (MgSO4), and
then concentrated under reduced pressure to give 43 mg of a
yellow oil. The mixture containing two major products was
separated by silica gel flash chromatography (3.5 × 6.5 cm)
eluting with 30% EtOAc/hexanes to give 2.8 mg (3.4 µmol,
6.4%) of tetraacetonide 2 (Rf ) 0.44, 50% EtOAc/hexanes) as
a yellow glass. The EtOAc concentration was raised to 65%
in 5% increments eluting 100 mL each time to isolate 11 mg
of a mixture of triacetonides (Rf ) 0.12, 50% EtOAc/hexanes).
The more polar material (24 mg) was then eluted with 100%
26-O-Acetyl-1′,3:5,7:9,11:13,15-tetr a k is-O-(1-m eth yleth -
ylid en e)-16,17,18,19,20,21,22,23,24,25-d eca h yd r ofilip in III
(5). Tetraacetonide 2 (2.0 mg, 2.4 µmol) was dissolved in 0.5
mL of THF and degassed with argon. Ac2O (30 µL, 0.32 mmol)
and DMAP (10 mg, 82 µmol) were added. The reaction was
allowed to stir under argon in the dark for 1.25 h. The reaction
was quenched with 50 µL of MeOH, and then the solution was
diluted with 20 mL of EtOAc. The organic solution was
washed with saturated NaHCO3, water, and brine, dried
(MgSO4), and then concentrated under reduced pressure to
give an oil. The product was purified by silica gel flash
chromatography (1 × 8 cm) eluting with 30% EtOAc/hexanes
to yield 1.0 mg of an oil. The product was further purified by
preparative reversed-phase HPLC eluting with 95% MeOH/
5% H2O at 2.5 mL/min. The major peak at 22.5 min was
collected to yield 0.3 mg of an oil. The oil was taken up in 1
mL of EtOAc. Pd(BaSO4) was added, and the solution was
stirred under H2 for 2.25 h. The catalyst was filtered and the
solvent evaporated under reduced pressure to yield 0.3 mg
(0.35 µmol, 14%) of a colorless oil: 1H NMR (500 MHz, C6D6)
δ 5.19-5.25 (m, 2 H), 4.49 (ddd, J ) 3.1, 7.7, 10.5 Hz, 1 H),
4.34-4.23 (m, 2 H), 4.22-4.09 (m, 2 H), 3.99 (m, 1 H), 3.89
(dd, J ) 7.2, 14.1 Hz, 1 H), 3.82 (m, 1 H), 2.59 (t, J ) 10.3 Hz,
1 H), 2.24-2.08 (m, 2 H), 2.00 (m, 1 H), 1.90-1.16 (m, 36 H),
1.82 (s, 3 H), 1.64 (s, 3 H), 1.58 (s, 3 H), 1.55 (s, 3 H), 1.41 (s,
3 H), 1.45 (s, 3 H), 1.44 (s, 3 H), 1.42 (s, 3 H), 1.38 (s, 3 H),
1.09 (d, J ) 6.3 Hz, 3 H), 1.04 (d, J ) 6.7 Hz, 3 H), 0.911 (t, J
) 7.1 Hz, 3 H); 13C NMR (125 MHz, C6D6, isotopically enriched
acetonide methyls) δ 30.8, 30.7, 30.1, 25.7, 24.8, 19.9 (2), 19.4;
12
HRMS (FAB) calcd for
C
13C2H83O12 853.5954, found 853.6042
46
(M - CH3).
1′,15,26-Tr i-O-a cetyl-3,5:7,9:11,13-tr is-O-(1-m eth yleth -
ylid en e)filip in III (3) a n d 1′,7,26-Tr i-O-a cetyl-3,5:9,11:13,-
15-tr is-O-(1-m eth yleth ylid en e)filip in III (4). The triace-
tonide mixture (13 mg, 16.8 µmol) from the preparation of
tetraacetonide 2 was taken up in 1 mL of THF and degassed
with argon. Ac2O (160 µL, 1.7 mmol) and DMAP (60 mg, 0.49
mmol) were added. The reaction was allowed to stir under
argon in the dark for 1.25 h. The reaction was quenched with
50 µL of MeOH, and then the solution was diluted with 40
mL of EtOAc. The organic solution was washed with saturated
NaHCO3, water, and brine, dried (MgSO4), and then concen-
trated under reduced pressure to give a yellow oil. The
mixture containing two products was separated by silica gel
flash chromatography (2 × 5 cm) eluting with 25% (200 mL)
and then 30% (200 mL) EtOAc/hexanes to give 9.5 mg (10.5
µmol, 64%) of triacetonide 3 (Rf ) 0.50, 50% EtOAc/hexanes)
and 2.9 mg (3.2 µmol, 19%) of triacetonide 4 (Rf ) 0.38, 50%
EtOAc/hexanes).
(32) Gen er a l Exp er im en ta l. Liquid chromatography was per-
formed using forced flow (flash chromatography) of the indicated
solvent system on EM reagent silica gel 60 (230-400 mesh). HPLC
purifications were carried out on
a C-18 reversed-phase column
(Spherisorb S5 ODS2, 250 × 10 mm) eluting with water/methanol
mixtures. Tetrahydrofuran and ether were distilled from benzophenone
ketyl. Methylene chloride was distilled from calcium hydride. Dim-
ethylformamide was distilled from barium hydroxide. Air- and/or
moisture-sensitive reactions were carried out under an atmosphere of
nitrogen or argon using oven-dried glassware and standard syringe/
septa techniques. NMR data for 13C DEPT experiments are reported
as quaternary (C), tertiary (CH), secondary (CH2), and primary (CH3)
carbon atoms. For overlapping signals the number of carbon atoms is
given in parentheses. GC analysis was carried out using an HP 5890
GC with an Alltech SE-54 column. Unless otherwise noted, the GC
temperature program was 50 °C for 5 min, then 50-250 °C at 5 °C/
min. A Chiraldex G TA capillary column (0.32 mm × 30 m) was used
to separate enantiomers.
Triacetonide 3: 1H NMR (500 MHz, C6D6) δ 6.51 (dd, J )
10.4, 13.6 Hz, 1 H), 6.45-6.30 (m, 4 H), 6.27-6.21 (m, 2 H),
6.13 (dd, J ) 10.6, 15.1 Hz, 1 H), 5.81 (dd, J ) 5.8, 15.6 Hz, 1
H), 5.69 (ddd, 3.0, 6.3, 10.0 Hz, 1 H), 5.62 (dd, J ) 4.6, 10.6
Hz, 1 H), 5.37 (dd, J ) 6.0, 9.3 Hz, 1 H), 5.31 (dq, J ) 6.0, 6.1

Filipin III
J . Org. Chem., Vol. 61, No. 13, 1996 4227
Hz, 1 H), 4.29 (ddd, J ) 2.1, 8.6, 10.8 Hz, 1 H), 4.12-4.07 (m,
1 H), 4.06-3.99 (m, 1 H), 3.90-3.81 (m, 2 H), 3.77-3.70 (m, 1
H), 3.11 (dd, J ) 6.1, 8.0 Hz, 1 H), 2.25 (ddd, J ) 3.4, 10.1,
13.5 Hz, 1 H), 2.10-1.16 (m, 19 H), 1.78 (s, 3 H), 1.73 (s, 3 H),
1.66, (s, 6 H), 1.55 (s, 3 H), 1.54 (s, 3 H), 1.45 (s, 3 H), 1.43 (s,
3 H), 1.30 (s, 3 H), 1.22 (s, 3 H), 1.15 (d, J ) 6.1 Hz, 3 H),
0.861 (t, J ) 6.9 Hz, 3 H); 13C NMR (125 MHz, C6D6,
isotopically enriched acetonide methyls) δ 30.8, 30.7, 30.0, 19.9,
19.8, 19.4; HRMS (FAB) calcd for C50H76O14 900.5237, found
900.5226 (M).
Triacetonide 4: 1H NMR (500 MHz, C6D6) δ 6.91 (dd, J )
11.5, 14.3 Hz, 1 H), 6.74 (dd, J ) 11.1, 14.5 Hz, 1 H), 6.66-
6.55 (m, 2 H), 6.50-6.34 (m, 3 H), 6.28 (dd, J ) 11.3, 14.1 Hz,
1 H), 6.09 (dd, J ) 4.9, 14.6 Hz, 1 H), 5.70 (ddd, J ) 2.8, 7.4,
10.1 Hz, 1 H), 5.49 (dd, J ) 5.0, 9.6 Hz, 1 H), 5.33 (dq, J )
5.0, 6.3 Hz, 1 H), 5.30 (m, 1 H), 4.52 (m, 1 H), 4.39 (d, J ) 5.6
Hz, 1 H), 4.28 (ddd, J ) 2.4, 7.3, 11.4 Hz, 1 H), 3.91-3.84 (m,
2 H), 3.74 (m, 1 H), 3.16 (dd, J ) 7.4, 7.4 Hz, 1 H), 2.63 (d, J
) 14.1 Hz, 1 H), 2.04-1.21 (m, 19 H), 1.87 (s, 3 H), 1.85 (s, 3
H), 1.77 (s, 3 H), 1.67 (s, 3 H), 1.56 (s, 3 H), 1.52 (s, 3 H), 1.43
(s, 3 H), 1.39 (s, 3 H), 1.33 (s, 3 H), 1.28 (s, 3 H), 1.19 (d, J )
6.3 Hz, 3 H), 0.840 (t, J ) 7.0 Hz, 3 H); 13C NMR (125 MHz,
C6D6, isotopically enriched acetonide methyls) δ 31.1, 30.5,
30.2, 24.7, 19.0, 19.0; HRMS (FAB) calcd for C50H76O14
900.5237, found 900.5150 (M).
PPTS (5 mg) was added. The solution was flushed with
nitrogen and allowed to stir for 9.75 h. The reaction was
quenched with 10 µL of Et3N and then the solvent evaporated
under reduced pressure to give a colorless oil. The desired
product was separated by silica gel chromatography. The
combined yield for the desired product was 32.3 mg (41.1 µmol,
15%); 1H NMR (500 MHz, C6D6) δ 5.07 (dq, J ) 6.2, 6.2 Hz, 1
H), 4.52 (ddd, J ) 2.1, 8.8, 11.1 Hz, 1 H), 4.36 (ddd, J ) 3.5,
6.0, 9.5 Hz, 1 H), 4.30 (ddd, J ) 2.9, 7.9, 10.8 Hz, 1 H), 4.12-
4.00 (m, 4 H), 3.88 (m, 1 H), 3.81 (ddd, J ) 1.3, 6.2, 8.2 Hz, 1
H), 3.75 (m, 1 H), 3.63 (m, 1 H), 3.49 (m, 1 H), 2.83 (dd, J )
8.6, 8.6 Hz, 1 H), 2.10-1.99 (m, 2 H), 1.71-1.15 (m, 19 H),
1.54 (s, 3 H), 1.52 (s, 3 H), 1.50 (s, 3 H), 1.45 (s, 3 H), 1.41 (s,
3 H), 1.36 (s, 3 H), 1.35 (s, 3 H), 1.343 (s, 3 H), 1.337 (s, 3 H),
1.26 (s, 3 H), 1.18 (d, J ) 5.6 Hz, 3 H), 1.14 (d, J ) 6.0 Hz, 3
H), 0.879 (t, J ) 6.2 Hz, 3 H); HRMS (FAB) calcd for C42H75O13
787.5209, found 787.5198 (M + H).
(S)-MTP A Ester of P en ta a ceton id e 7. Pentaacetonide
7 (2.8 mg, 3.6 µmol) was dissolved in 2 mL of CH2Cl2. DMAP
(4.0 mg, 33 µmol) and (R)-MTPA-Cl (3 µL, 16 µmol) were
added. The reaction was allowed to stir under N2 for 1 h. The
reaction mixture was diluted with 10 mL of CH2Cl2 and
washed with saturated NaHCO3. The aqueous layer was
extracted (2 × 10 mL) with CH2Cl2. The combined organic
layers were washed with water and brine, dried (Na2SO4), and
then concentrated under reduced pressure to give a colorless
oil. The product was purified by silica gel flash chromatog-
raphy (1 × 5 cm) eluting with 30% EtOAc/hexanes to yield
3.6 mg (3.6 µmol, 100%) of a yellow oil: 1H NMR (500 MHz,
C6D6) δ 7.48-7.52 (m, 2 H), 7.37-7.40 (m, 3 H), 5.45 (ddd, J
) 5.5, 5.6, 8.3 Hz, 1 H), 4.80 (dq, J ) 6.1, 6.2 Hz, 1 H), 4.23
(m, 1 H), 4.06-3.89 (m, 7 H), 3.84 (ddd J ) 3.0, 6.9, 10.5 Hz,
1 H), 3.71-3.65 (m, 2 H), 3.53 (s, 3 H), 2.91 (dd, J ) 7.0, 8.0
Hz, 1 H), 1.82-1.76 (m, 2 H), 1.68-1.60 (m, 2 H), 1.48-1.10
(m, 19 H), 1.44 (s, 3 H), 1.43 (s, 3 H), 1.42 (s, 3 H), 1.41 (s, 3
H), 1.37 (s, 3 H), 1.36 (s, 6 H), 1.32 (s, 3 H), 1.31 (s, 3 H), 1.28
(s, 3 H), 1.20 (d, J ) 6.4 Hz, 3 H), 0.846 (t, J ) 7.1 Hz, 3 H);
HRMS (FAB) calcd for C52H82O15F3 1003.5607, found 1003.5541
(M + H).
[2(1R),3S,5R,7S,9S,11R,13S,15S]-2-(1-(Acetyloxy)h exyl)-
7-O-a ce t yl-3,5:9,11:13,15-t r is-O-(1-m e t h yle t h ylid e n e )-
3,5,7,9,11,13,15-h epta h yd r oxy-16-oxoh ep ta d eca n oic Acid ,
(2R,3R)-2-(Acetyloxy)-1-m eth yl-3-oxopr opyl Ester (6). Tri-
acetonide 4 (1.5 mg, 1.7 µmol) was dissolved in 1 mL of CH2-
Cl2. MeOH (100 µL) was added and the solution cooled to -78
°C. Ozone was bubbled through the solution until a blue color
persisted, and then nitrogen was bubbled through until the
solution was clear. PPh3 (4 mg, 15 µmol) was added and the
reaction mixture allowed to warm to room temperature. After
overnight stirring, the reaction mixture was concentrated
under reduced pressure to give a colorless oil. The product
was purified by silica gel flash chromatography (1 × 5 cm)
eluting with 60% EtOAc/hexanes to yield 0.6 mg (0.7 µmol,
1
(R)-MTP A Ester of P en ta a ceton id e 7. The (R)-MTPA
ester of pentaacetonide 7 was prepared from pentaacetonide
7 following the same procedure used to prepare the (S)-MTPA
ester described above. 7: yield 3.5 mg, 3.5 µmol, 97%; 1H NMR
(500 MHz, C6D6) δ 7.54-7.50 (m, 2 H), 7.36-7.41 (m, 3 H),
5.47 (dd, J ) 6.3, 12.4 Hz, 1 H), 4.85 (dq, J ) 6.3, 6.3 Hz, 1
H), 4.23 (m, 1 H), 4.04-3.89 (m, 8 H), 3.71 (ddd, J ) 4.2, 8.7,
11.6 Hz, 1 H), 3.67 (m, 1 H), 3.47 (s, 3 H), 2.93 (dd, J ) 7.3,
7.3 Hz, 1 H), 1.81-1.74 (m, 2 H), 1.65-1.60 (m, 2 H), 1.42 (s,
3 H), 1.41 (s, 3 H), 1.39 (s, 3 H), 1.36 (s, 3 H), 1.35 (s, 3 H),
1.33 (s, 3 H), 1.32 (s, 6 H), 1.31 (s, 3 H), 1.30 (s, 3 H), 1.20 (d,
J ) 6.4 Hz, 1 H), 0.821 (t, J ) 6.8 Hz, 1 H); HRMS (FAB)
calcd for C52H82O15F3 1003.5607, found 1003.5659 (M + H).
(2S,3R)-1,2-Di-O-isop r op ylid en e-1,2,3-bu ta n etr iol (8).
Pentaacetonide 7 (22 mg, 28 µmol) was placed in a flask and
flushed with argon. In another flask LAH (32 mg, 840 µmol)
was dissolved in 5 mL of THF and then transferred via cannula
to the reaction flask. After the mixture was stirred for 1.75
h, the reaction was quenched with Na2SO4‚10H2O (200 mg).
The solids were filtered and washed with THF. The organic
solution was concentrated under reduced pressure to give a
colorless oil. The desired product was separated from the
mixture by silica gel flash chromatography (2 × 5 cm) eluting
with 70% Et2O/pentane to give acetonide alcohol 8 (4.1 mg,
28 µmol, 100%) as a colorless oil. A mixture of acetonide
isomers (14.2 mg, 22 µmol, 79%) representing the large polyol
was also collected.
41%) of a colorless oil; H NMR (500 MHz, C6D6) δ 9.18 (s, 1
H), 5.66 (ddd, J ) 2.8, 5.8, 9.2 Hz, 1 H), 5.45 (m, 1 H), 5.38
(dq, J ) 3.4, 6.7 Hz, 1 H), 5.06 (d, J ) 3.4 Hz, 1 H), 4.34 (dd,
J ) 10.5, 11.1 Hz, 1 H), 4.12-3.89 (m, 5 H), 3.17 (dd, J ) 7.6,
7.6 Hz, 1 H), 2.04-1.10 (m, 23 H), 1.78 (s, 3 H), 1.77 (s, 3 H),
1.72 (s, 3 H), 1.49 (s, 3 H), 1.42 (s, 3 H), 1.32 (s, 3 H), 1.31 (s,
6 H), 1.29 (s, 3 H), 1.12 (d, J ) 6.6 Hz, 3 H), 0.855 (t, J ) 6.5
Hz, 3 H); 13C NMR (125 MHz, C6D6, isotopically enriched
acetonide methyls) δ 30.5, 30.3, 25.9, 24.3, 19.6, 19.3.
[2(1R),3S,5R,7S,9R,11S,13S,15S,16S]- a n d [2(1R),3S,-
5R,7S,9R,11S,13S,15S,16R]-2-(1-H yd r oxyh exyl)-3,5:7,9:
11,13:15,16-tetr a k is-O-(1-m eth yleth ylid en e)-3,5,7,9,11,13,-
15,16-octa h yd r oxyh ep ta d eca n oic Acid , [1S,1(4R)]-1-(2,2-
Dim et h yl-1,3-d ioxola n -4-yl)et h yl E st er (7). Per-TES-
filipin (470 mg, 280 µmol) was dissolved in 20 mL of a 2:1
mixture of CH2Cl2/MeOH and cooled to -78 °C. Ozone was
bubbled through the solution until a blue color persisted, and
then nitrogen was bubbled through until it was clear. NaBH4
(200 mg, 5.3 mmol) was added and the reaction mixture
allowed to warm to room temperature. After the mixture was
stirred for 2 h, the reaction was quenched with 10 mL of
saturated NaHCO3. The aqueous phase was extracted (3 ×
10 mL) with CH2Cl2. The combined organic portions were
washed with water and brine, dried (Na2SO4), and then
concentrated under reduced pressure to give 430 mg of a
colorless oil. The colorless oil was taken up in 20 mL of acetone
and treated with Dowex 50w × 1 ion exchange resin (H+ form).
After being stirred for 2.5 h, the reaction mixture was filtered
and concentrated under reduced pressure to give a yellow oil.
The oil was taken up in 60 mL of Et2O, washed with saturated
NaHCO3, water and brine, dried (MgSO4), and then concen-
trated under reduced pressure to give a colorless oil. The
desired product was separated from the mixture by silica gel
flash chromatography (4 × 8 cm) eluting with 30% EtOAc/
hexanes to give 15 mg of a colorless oil. The more polar
material was eluted with 100% EtOAc. The more polar
material was taken up in 10 mL of acetone and 5 mL of DMP.
Acetonide alcohol 8: 1H NMR (300 MHz, CDCl3) δ 4.03-
3.87 (m, 4 H), 1.93 (d (broad), J ) 2.6 Hz, 1 H), 1.42 (s, 3 H),
1.35 (s, 3 H), 1.14 (d, J ) 6.4 Hz, 3 H); 13C NMR (75 MHz,
CDCl3) δ 109.1, 79.4, 66.7, 64.4, 26.4, 25.2, 18.2; HRMS (CI)
calcd for C7H15O3 147.1021, found 147.1017 (M + H).
(S)-MTP A a n d (R)-MTP A Ester s of (2S,3R)-1,2-Di-O-
isop r op ylid en e-1,2,3-bu ta n etr iol (8). Acetonide alcohol 8
(1.5 mg, 10 µmol) was dissolved in 2 mL of CH2Cl2. DMAP
(10 mg, 82 µmol) and (R)-MTPA-Cl (8 µL, 43 µmol) were added.
The reaction mixture was allowed to stir under N2 for 45 min,

4228 J . Org. Chem., Vol. 61, No. 13, 1996
Richardson and Rychnovsky
1
diluted with 10 mL of CH2Cl2, and washed with saturated
NaHCO3. The aqueous layer was extracted (2 × 10 mL) with
CH2Cl2. The combined organic layers were washed with water
and brine, dried (Na2SO4), and then concentrated under
reduced pressure to give a colorless oil. The product was
purified by silica gel flash chromatography (1 × 5 cm) eluting
with 30% EtOAc/hexanes to yield 3.2 mg (8.8 µmol, 88%) of a
yellow oil: 1H NMR (500 MHz, CDCl3) δ 7.48-7.52 (m, 2 H),
7.37-7.40 (m, 3 H), 5.07 (dq, J ) 6.4, 6.4 Hz, 1 H), 4.01 (ddd,
J ) 6.4, 6.3, 6.1 Hz, 1 H), 3.87 (dd, J ) 6.3, 8.6 Hz, 1 H), 3.61
(dd, J ) 6.1, 8.6 Hz, 1 H), 3.56 (s, 3 H), 1.39 (d, J ) 6.4 Hz, 3
H), 1.31 (s, 3 H), 1.29 (s, 3 H); HRMS (CI) calcd for C17H22O5F3
363.1420, found 363.1415 (M + H).
1137, 1060, 991 cm-1; H NMR (300 MHz, CDCl3) δ 3.85 (dt,
J ) 11.1, 6.4 Hz, 2 H), 6.67-3.52 (m, 3 H), 3.20 (broad s, 2 H),
1.26 (d, J ) 6.3, 3 H); 13C NMR (75 MHz, CDCl3, DEPT) δ
CH, 75.3, 68.0; CH2, 46.6; CH3, 19.4; HRMS (CI) calcd for
C4H13ClNO2 142.0636, found 142.0640 (M + NH4). Anal.
Calcd for C4H9ClO2: C, 38.57; H, 7.28. Found: C, 38.32; H,
7.19.
(4R ,5S )-4-(Ch lor om e t h yl)-2,2,5-t r im e t h yl-1,3-d ioxo-
la n e (22). Chloro diol 21 (1.25 g, 10 mmol) was dissolved in
60 mL of acetone and 30 mL of DMP. PPTS (260 mg, 1 mmol)
was added. The reaction mixture was tightly capped under
N2 and stirred for 22 h. The reaction was quenched with 250
µL Et3N, and then the mixture was concentrated under
reduced pressure. The crude product was purified by silica
gel flash chromatography (4 × 8 cm) eluting with 5% Et2O/
pentane to yield 1.22 g (7.46 mmol, 74%) of product as an 86:
14 mixture of anti (tR ) 17.11 min) to syn (tR ) 18.18 min)
isomers by GC analysis: IR (neat) 2987, 2936, 2883, 1380,
1246, 1221, 1173, 1090, 1048 cm-1; 1H NMR (300 MHz, CDCl3)
δ 3.98 (dt, J ) 7.6, 6.0 Hz, 1 H), 3.75 (m, 1 H), 3.60 (dd, J )
4.8, 11.4 Hz, 1 H), 3.54 (dd, J ) 5.8, 11.4 Hz, 1 H), 1.41 (s, 3
H), 1.38 (s, 3 H), 1.33 (d, J ) 6.0 Hz, 3 H); 13C NMR (75 MHz,
CDCl3, DEPT) δ C, 109.0; CH, 81.5, 75.4; CH2, 43.8; CH3, 27.4,
26.8, 18.4; HRMS (CI) calcd for C7H14ClO2 165.0683, found
165.0679 (M + H). Anal. Calcd for C7H13ClO2: C, 51.07; H,
7.96. Found: C, 50.97; H, 7.83.
The (R)-MTPA ester of acetonide alcohol 8 was prepared
from acetonide alcohol 8 following the same procedure used
to prepare the (S)-MTPA ester described above. 8: yield 3.4
1
mg, 9.4 µmol, 94%; H NMR (500 MHz, CDCl3) δ 7.50-7.52
(m, 2 H), 7.37-7.40 (m, 3 H), 5.14 (dq, J ) 6.2, 6.3 Hz, 1 H),
4.10 (ddd, J ) 6.4, 6.4, 6.2 Hz, 1 H), 3.99 (dd, J ) 6.4, 8.4 Hz,
1 H), 3.73 (dd, J ) 6.4, 8.4 Hz, 1 H), 3.51 (s, 3 H), 1.36 (s, 3
H), 1.33 (s, 3 H), 1.29 (d, J ) 6.3 Hz, 3 H); HRMS (CI) calcd
for C17H22O5F3 363.1420, found 363.1432 (M + H).
[1(4R,5R),1S,3R,5S,7R,9S,11R,13S,14S]- a n d [1(4R,5R),-
1S,3R,5S,7R,9S,11R,13S,14R]-1-(2,2-Dim et h yl-4-p en t yl-
1,3-d ioxa n -5-yl)-1,3:5,7:9,11:13,14-t et r a k is-O-(1-m et h yl-
eth ylid en e)p en ta d eca n -1,3,5,7,9,11,13,14-octol (9). The
mixture of acetonide isomers from the preparation of acetonide
alcohol 8 was dissolved in 5 mL of acetone and 2.5 mL of DMP.
PPTS (5 mg) was added, and the reaction was stirred for 11
h. The reaction was quenched with 5 µmol Et3N, and then
the solution was concentrated under reduced pressure to give
a colorless oil. The product was purified by silica gel flash
chromatography (3 × 5 cm) eluting with 30% EtOAc/hexanes
to give pentaacetonide 9 (14.3 mg, 20.9 µmol, 96%) as an
inseparable 1:1 mixture of C16 epimers. The product was a
Analysis of the product by GC on a chiral column (Chiraldex
G TA) showed the major isomer was 81% ee and the minor
isomer was 47% ee.
(4R,5S)-4-(Iod om et h yl)-2,2,5-t r im et h yl-1,3-d ioxola n e
(23). Acetonide 22 (1.10 g, 6.72 mmol) was placed in a 100
mL flask with 40 mL of xylenes. KI (27.9 g, 168 mmol) and
18-crown-6 (2.66 g, 10.0 mmol) were added. The reaction was
heated to reflux and stirred for 4 days under N2. The reaction
was quenched with 20 mL of saturated Na2SO3, and then the
solution was diluted with water until all the solids dissolved.
The aqueous layer was extracted (3 × 25 mL) with CH2Cl2.
The combined organic layers were washed with water and
brine, dried (Na2SO4), and then concentrated under reduced
pressure. The crude product was purified by silica gel flash
chromatography (6 × 6 cm) eluting with 5% Et2O/pentane to
yield 1.51g of a colorless oil that was a 35:65 mixture of
starting material as a 2:1 mixture of anti (tR ) 17.13 min) to
syn (tR ) 18.19 min) and product as a 97:3 mixture of anti (tR
) 23.66 min) to syn (tR ) 24.76 min) by GC analysis. The
product was a colorless oil: IR (neat) 2985, 2934, 2880, 1379,
colorless oil: [R]26 -20.5° (c ) 1.43, CHCl3); 13C NMR (125
D
MHz, C6D6, isotopically enriched acetonide methyls) δ 29.0,
27.8, 27.6, 26.1 (1,2 diol acetonides) 30.8, 30.7, 30.7, 30.2, 20.0,
19.9, 19.8, 19.0 (1,3-diol acetonides); HRMS (FAB) calcd for
C38H69O10 685.4893, found 685.4929 (M + H).
[13S,14S] isom er : 1H NMR (500 MHz, CDCl3) δ 4.28-4.19
(m, 1 H), 4.06-3.95 (m, 7 H), 3.81 (dd, J ) 3.2, 12.1 Hz, 1H),
3.68-3.66 (m, 2 H), 1.84-1.76 (m, 2 H), 1.72-1.10 (m, 19 H),
1.423 (s, 3H), 1.41 (s, 6H), 1.40 (s, 3 H), 1.36 (s, 3 H) 1.35 (s,
6 H), 1.34 (s, 3 H), 1.32 (s, 3 H), 1.31 (s, 3 H), 1.23 (d, J ) 5.4
Hz, 3 H), 0.866 (t, J ) 6.6 Hz, 3 H); 13C NMR (125 MHz, CDCl3)
δ 107.7, 98.5, 98.4, 98.3 (2), 79.0, 77.1, 71.7, 66.7 (2), 65.7, 65.2
(2), 65.0, 61.0, 42.7 (2), 41.1, 39.6, 37.5, 36.5, 34.8, 33.1, 31.9,
30.3, 30.2 (2), 29.7, 27.3 (2), 25.3, 22.6, 20.0, 19.8 (2), 18.9,
17.2, 14.1.
1243, 1222, 1175, 1095 cm-1
3.85 (dt, J ) 13.5, 6.1 Hz, 1 H), 3.55 (m, 1 H), 3.24-3.22 (m,
2 H), 1.41 (s, 3 H), 1.40 (s, 3 H), 1.33 (d, J ) 6.0 Hz, 3 H); 13
NMR (75 MHz, CDCl3, DEPT) δ C, 108.7; CH, 81.1, 77.4; CH2,
4.9; CH3, 27.6, 27.2, 18.6; HRMS (CI) calcd for C7H14IO2
257.0039, found 257.0048 (M + H).
;
1H NMR (300 MHz, CDCl3) δ
C
[13S,14R] isom er : 1H NMR (500 MHz, CDCl3) δ 4.28-4.19
(m, 1 H), 4.06-3.95 (m, 7 H), 3.78 (dd, J ) 3.2, 12.1 Hz, 1 H),
3.64-3.56 (m, 2 H), 1.84-1.76 (m, 2 H), 1.72-1.10 (m, 19 H),
1.42 (s, 3 H), 1.41 (s, 3 H), 1.40 (s, 3 H), 1.39 (s, 3 H), 1.36 (s,
6 H), 1.35 (s, 3 H), 1.34 (s, 3 H), 1.32 (s, 3 H), 1.30 (s, 3 H),
1.11 (d, J ) 6.1 Hz, 3 H), 0.866 (t, J ) 6.6 Hz, 3 H); 13C NMR
(125 MHz, CDCl3) δ 107.2, 98.8, 98.5, 98.4, 98.3, 77.1, 74.4,
73.8, 70.3, 68.7, 66.3, 65.2 (2), 65.0, 59.0, 45.1, 42.7, 42.5, 41.1,
37.5, 36.9, 34.7, 31.7, 30.2 (2), 30.1, 29.7, 28.6, 25.9, 24.8, 22.6,
21.2, 20.0, 19.8 (2), 18.9, 17.1, 15.8.
Syn th esis of Cor r ela tion F r a gm en t 10. (2R,3S)-1-
Ch lor obu ta n e-2,3-d iol (21). AD-mix-R (17.0 g, 12 mmol),
NaHCO3 (3.05 g, 36 mmol), methanesulfonamide (1.15 g, 12
mmol), 60 mL of water, and 60 mL of tert-butyl alcohol were
placed in a 500 mL flask equipped with an overhead stirrer.
The solution was stirred at room temperature until all the
solids dissolved. The solution was cooled to 0 °C, and then
crotyl chloride (1.09 g, 12 mmol) was added. The solution was
stirred for 3 days. The reaction was quenched with 18 g of
Na2SO3, and the solution was stirred for 1 h at room temper-
ature. The aqueous layer was extracted (3 × 90 mL) with
EtOAc. The combined organic layers were dried (MgSO4) and
then concentrated under reduced pressure. The crude product
was purified by silica gel flash chromatography (4 × 6 cm)
eluting with 90% Et2O/pentane to yield 1.29 g (10.3 mmol,
86%) of a colorless oil: IR (neat) 3372, 2976, 2934, 2911, 1321,
(2S,3S,5S,7S)-2,3:5,7-Bis-O-(1-m eth yleth yliden e)-8-ch lo-
r o-5-cya n oocta n e-2,3,5,7-tetr ol (24). A solution of chloro
cyanohydrin acetonide 13 (760 mg, 4.0 mmol) in 2 mL of THF
was cooled to -78 °C and then added via cannula to 4.7 mmol
of LiNEt2 in 5 mL of THF under Ar at -78 °C. The reaction
was stirred for 1 h. A solution of compound 23 (512 mg, 1.49
mmol) in 2 mL of THF was cooled to -78 °C and then
transferred to the reaction flask via cannula. The solution was
stirred for 1 h and then warmed to -27 °C with a MeOH/ice
bath. The reaction was allowed to warm slowly to 15 °C over
15 h. The reaction was quenched with 6 mL of saturated NH4-
Cl. Water was added until all the solids dissolved. The
solution was diluted with 25 mL of CH2Cl2. The aqueous layer
was extracted (3 × 25 mL) with CH2Cl2. The combined organic
layers were washed with brine, dried (Na2SO4), and then
concentrated under reduced pressure. The crude product was
purified by silica gel flash chromatography (3 × 8 cm) eluting
with 10% EtOAc/hexanes to yield 361.4 mg (1.14 mmol, 76%)
of a colorless oil which crystallized upon standing: mp ) 46-
48 °C; IR (neat) 2987, 2936, 2886, 1382, 1251, 1206, 1177, 1102
cm-1; H NMR (500 MHz, CDCl3) δ 4.34 (ddt, J ) 1.8, 11.7,
5.3 Hz, 1 H), 3.82 (dt, J ) 3.2, 8.2 Hz, 1 H), 3.74 (m, 1 H), 3.55
(dd, J ) 5.5, 11.4 Hz, 1 H), 3.50 (dd, J ) 5.2, 11.4 Hz, 1 H),
2.18 (dd, J ) 2.1, 13.9 Hz, 1 H), 2.02 (dd, J ) 3.2, 14.2 Hz, 1
1

Filipin III
J . Org. Chem., Vol. 61, No. 13, 1996 4229
H), 1.95 (dd, J ) 8.1, 14.2 Hz, 1 H), 1.72 (s, 3 H), 1.62 (dd, J
) 11.7, 13.9 Hz, 1 H), 1.39 (s, 3 H), 1.38 (s, 3 H), 1.35 (s, 3 H),
1.28 (d, J ) 6.0 Hz, 3 H); 13C NMR (75 MHz, CDCl3, DEPT) δ
C, 120.9, 108.7, 101.6, 69.1; CH, 77.8, 77.0, 66.7; CH2, 46.3,
45.2, 37.6; CH3, 30.7, 27.3, 27.1, 21.6, 17.0; HRMS (CI) calcd
for C15H25ClNO4 318.1473, found 318.1484 (M + H). Anal.
Calcd for C15H24ClNO4: C, 56.69; H, 7.61. Found: C, 56.50;
H, 7.75.
(27). A 25 mL Schlenk flask was charged with triacetonide
dinitrile 26 (250 mg, 0.531 mmol), 18-crown-6 (168 mg, 0.636
mmol), and 5 mL of xylenes. The solution was degassed with
Ar for 15 min followed by addition of KI (5.7 g, 34 mmol). The
reaction was heated to reflux and stirred for 2 days. The
reaction mixture was quenched with 25 mL of saturated Na2-
SO3. The aqueous layer was extracted (3 × 25 mL) with CH2-
Cl2. The combined organic layers were washed with water and
brine, dried (Na2SO4), and then concentrated under reduced
pressure. The crude product was purified by silica gel flash
chromatography eluting with 10-20% EtOAc/hexane to yield
280 mg (0.498 mmol, 94%) of a colorless oil: [R]24D ) 38.3 (c )
1.07, CHCl3); IR (neat) 2983, 2934, 2875, 1379 cm-1; 1H NMR
(500 MHz, CDCl3) δ 4.50 (dddd, J ) 1.8, 4.1, 6.3, 11.6 Hz, 1
H), 4.02 (dddd, J ) 2.3, 5.2, 5.2, 11.0 Hz, 1 H), 3.80 (dt, J )
3.1, 8.2 Hz, 1 H), 3.73 (dq, J ) 8.2, 6.0 Hz, 1 H), 2.23 (dd, J )
5.2, 10.6 Hz, 1 H), 3.19 (dd, J ) 5.2, 10.6 Hz, 1 H), 2.97-2.08
(m, 5 H), 1.92 (dd, J ) 8.2, 14.1 Hz, 1 H), 1.89 (dd, J ) 2.3,
13.6 Hz, 1 H), 1.74 (s, 3 H), 1.70 (s, 3 H), 1.61 (dd, J ) 13.6,
11.6 Hz, 1 H), 1.44 (s, 3 H), 1.37 (s, 3 H), 1.36 (s, 3 H), 1.34 (s,
3 H), 1.27 (d, J ) 6.0 Hz, 3 H); 13C NMR (75 MHz, CDCl3,
DEPT) δ C, 121.7, 120.9, 108.7, 101.9, 101.1, 65.5, 68.3; CH,
77.8, 76.9, 65.8, 61.7; CH2, 46.0, 45.1, 39.9, 38.0, 8.6; CH3, 30.8,
(2S,3S,5S,7S)-2,3:5,7-Bis-O-(1-m eth yleth ylid en e)-5-cy-
a n o-8-iod oocta n e-2,3,5,7-tetr ol (25). A 25 mL Schlenk flask
was charged with diacetonide nitrile 24 (437 mg, 1.37 mmol),
18-crown-6 (500 mg, 1.89 mmol), and 10 mL of xylenes. The
solution was degassed with Ar for 15 min followed by addition
of KI (5.7 g, 34 mmol). The reaction was heated to reflux and
stirred for 2 days. The reaction was quenched with 20 mL of
saturated Na2SO3, and then the solution was diluted with
water until all the solids dissolved. The aqueous layer was
extracted (3 × 25 mL) with CH2Cl2. The combined organic
layers were washed with water and brine, dried (Na2SO4), and
then concentrated under reduced pressure. The crude product
was purified by silica gel flash chromatography eluting with
10% EtOAc/hexane to yield 555 mg (1.36 mmol, 99%) of a
colorless oil: [R]26 ) 22.3 (c ) 0.96, CHCl3); IR (neat) 3032,
D
30.6, 27.3, 27.1, 21.7, 21.5, 17.0; HRMS (FAB) calcd for C23H36
-
2985, 2970, 2933, 2886, 1430, 1380, 1298, 1265, 1236, 1179,
1136, 1100, 1074, 1047, 1036, 1012, 984, 948, 936 cm-1 1H
;
IN2O6 563.1620, found 563.1625 (M + H). Anal. Calcd for
C23H35IN2O6: C, 49.12; H, 6.27. Found: C, 49.12; H, 6.36.
(Z)-2-Octen -1-ol (29). A sample of Ni(OAc)2‚4H2O (4.00
g, 16.0 mmol) was placed in a Parr hydrogenation bottle with
150 mL of absolute EtOH. The solution was purged with a
stream of N2. A solution of NaBH4 (600 mg, 16.0 mmol) in 20
mL of EtOH was transferred to the hydrogenation bottle via
cannula. The solution was stirred vigorously for 15 min.
Ethylenediamine (2.0 mL, 32 mmol) was added via syringe. A
sample of 2-octyn-1-ol (20.0 g, 160 mmol) was transferred to
the hydrogenation bottle via cannula. The solution was
shaken in a Parr hydrogenation apparatus under 50 psi of H2
for 2 h. The catalyst was filtered off through charcoal. The
purple solution was diluted with 300 mL of water and 200 mL
of Et2O. The layers were separated, and then the aqueous
layer was extracted (3 × 100 mL) with Et2O. The combined
organic layers were washed with water and brine, dried
(MgSO4), and then concentrated under reduced pressure to
give 21 g of a slightly yellow oil. Distillation (40 °C at 0.4
Torr) gave 18.9 g (147 mmol, 92%) of a colorless oil: IR (neat)
3327, 3016, 2958, 2937, 2867, 1656, 1453, 1380, 1340, 1312,
NMR (500 MHz, CDCl3) δ 4.09 (dddd, J ) 2.0, 5.1, 6.3, 11.5
Hz, 1 H), 3.79 (ddd, J ) 3.3, 8.1, 8.1 Hz, 1 H), 3.72 (dq, J )
8.1, 6.0 Hz, 1 H), 3.18 (dd, J ) 5.1, 10.4 Hz, 1 H), 3.13 (dd, J
) 6.3, 10.4 Hz, 1 H), 2.22 (dd, J ) 2.0, 13.8 Hz, 1 H), 1.99 (dd,
J ) 3.3, 14.2 Hz, 1 H), 1.93 (dd, J ) 8.1, 14.2 Hz, 1 H), 1.68 (s,
3 H), 1.49 (dd, J ) 11.5, 13.8 Hz, 1 H), 1.37 (s, 3 H), 1.35 (s, 3
H), 1.32 (s, 3 H), 1.26 (d, J ) 6.0 Hz, 3 H); 13C NMR (75 MHz,
CDCl3, DEPT) δ C, 120.9, 108.7, 101.9, 69.1; CH, 77.8, 76.9,
66.4; CH2, 45.1, 40.1, 7.6; CH3, 30.7, 27.3, 27.1, 21.6, 17.1;
HRMS (CI) calcd for C15H28IN2O2 427.1094, found 427.1099
(M + NH4). Anal. Calcd for C15H24INO4: C, 44.02; H, 5.91.
Found: C, 44.22; H, 5.84.
(2S,3S,5S,7R,9S,11S)-5,9-Dicya n o-12-ch lor o-2,3:5,7:9,11-
t r is-O-(1-m et h ylet h ylid en e)d od eca n e-2,3,5,7,9,11-h exol
(26). A solution of chloro cyanohydrin acetonide 13 (130 mg,
0.68 mmol) in 300 µL of THF was added via syringe to 0.81
mmol of LiNEt2 in 1 mL of THF under Ar at -63 °C. The
reaction mixture was stirred for 1 h, then cooled to -78 °C,
and stirred for another hour. DMPU (400 µL, 3.31 mmol) and
a solution of iodide 25 (140 mg, 0.34 mmol) in 250 µL of THF
were added to the reaction flask via syringe. The solution was
stirred for 15 min and then warmed to -25 °C with a MeOH/
ice bath. The reaction mixture was allowed to warm slowly
to 15 °C over 16 h. The reaction was quenched with 5 mL of
saturated NH4Cl. Water was added until all the solids
dissolved. The aqueous layer was extracted (3 × 20 mL) with
CH2Cl2. The combined organic layers were washed with brine,
dried (Na2SO4), and then concentrated under reduced pressure.
The crude product was purified by silica gel flash chromatog-
raphy (4 × 10 cm) eluting with 10% EtOAc/hexanes to yield
151 mg (0.32 mmol, 94%) of a colorless oil which crystallized
upon standing. Recrystallization from hexanes gave 105.8 mg
(0.22 mmol, 66%) of a white crystalline solid: mp ) 123-124°
C; [R]26D ) 39.9° (c ) 1.30, CHCl3); IR (KBr) 2986, 2939, 2988,
1
1022 cm-1; H NMR (300 MHz, CDCl3) δ 5.60-5.45 (m, 2 H),
4.15 (t, J ) 5.6 Hz, 2 H), 2.03 (q, J ) 6.8 Hz, 2 H), 1.82 (d, J
) 5.2 Hz, 1 H), 1.38-1.18 (m, 6 H), 0.857 (t, J ) 6.6 Hz, 3 H);
13C NMR (75 MHz, CDCl3, DEPT) δ CH, 133.3, 128.3; CH2,
58.6, 31.4, 29.3, 27.4, 22.5; CH3, 14.0; HRMS (EI) calcd for
C8H16O 128.1202, found 128.1199 (M). Anal. Calcd for
C8H16O: C, 74.94; H, 12.58. Found: C, 74.93; H, 12.48.
(2S,3R)-3-P en tyloxir a n e-2-m eth a n ol (30). (E)-2-Octen-
1-ol, Ti(OiPr)4, and L-(+)-diethyl tartrate were freshly distilled.
A 500 mL two-neck flask was flame dried and flushed with
Ar. Activated 4 Å sieves (1.5 g) and 200 mL of CH2Cl2 were
added. The solution was cooled to -20 °C. A solution of L-(+)-
diethyl tartrate (1.54 g, 7.50 mmol) in 10 mL of CH2Cl2, a
solution of Ti(OiPr)4 (1.86 mL, 6.2 mmol) in 10 mL of CH2Cl2,
and tert-butyl hydroperoxide (13.4 mL of a 5.8 M solution in
decane) were each separately stirred with 4 Å sieves for 15
min and then transferred to the reaction flask via cannula.
The reaction was stirred at -20 °C for 40 min. Then a solution
of (Z)-2-octen-1-ol (4.0 g, 31.2 mmol) in 10 mL of CH2Cl2 was
stirred with 4 Å sieves for 15 min and then transferred to the
reaction flask via cannula. The reaction mixture was stirred
at -15 °C for 45 h. The reaction mixture was then warmed
to 0 °C, 35 mL of water was added, and the mixture was stirred
for 1 h. A 30% NaOH aqueous solution saturated with NaCl
(6 mL) was added. After 20 min of stirring, the solution had
separated into two phases. The aqueous phase was extracted
(2 × 100 mL) with CH2Cl2. The combined organic layers were
washed with brine. The brine was back extracted (2 × 30 mL)
with CH2Cl2. The combined organic solutions were dried (Na2-
SO4) and then concentrated under reduced pressure. The
crude product was purified by silica gel flash chromatography
1
2879, 1384 cm-1; H NMR (500 MHz, CDCl3) δ 4.49 (dddd, J
) 2.0, 3.5, 7.3, 11.7 Hz, 1 H), 4.35 (dddd, J ) 2.1, 4.9, 5.1,
11.8 Hz, 1 H), 3.80 (dt, J ) 2.9, 8.3 Hz, 1 H), 3.73 (dq, J ) 8.2,
5.9 Hz, 1 H), 3.57 (dd, J ) 4.9, 11.5 Hz, 1 H), 3.52 (dd, J )
5.1, 11.5 Hz, 1 H), 2.11-2.03 (m, 4 H), 2.00 (dd, J ) 2.9, 14.1
Hz, 1 H), 1.92 (dd, J ) 8.4, 14.1 Hz, 1 H), 1.86 (dd, J ) 2.1,
13.6 Hz, 1 H), 1.72 (s, 3 H), 1.71 (s, 3 H), 1.62 (dd, J ) 11.8,
14.0 Hz, 1 H), 1.43 (s, 3 H), 1.364 (s, 3 H), 1.357 (s, 3 H), 1.34
(s, 3 H), 1.28 (d, J ) 5.9 Hz, 3 H); 13C NMR (75 MHz, CDCl3,
DEPT) δ C, 121.6, 121.0, 108.7, 101.7, 101.2, 69.5, 67.6; CH,
77.8, 76.9, 66.4, 61.8; CH2, 46.5, 46.1, 54.1, 39.9, 35.2; CH3,
30.8, 30.6, 27.2, 27.1, 21.5, 21.4, 17.0; HRMS (CI) calcd for
C23H39ClN3O6 488.2527, found 488.2536 (M + NH4). Anal.
Calcd for C23H35ClN2O6: C, 58.65; H, 7.49. Found: C, 58.47;
H, 7.40.
(2S,3S,5S,7R,9S,11S)-5,9-Dicya n o-12-iod o-2,3:5,7:9,11-
t r is-O-(1-m et h ylet h ylid en e)d od eca n e-2,3,5,7,9,11-h exol

4230 J . Org. Chem., Vol. 61, No. 13, 1996
Richardson and Rychnovsky
(6 × 5 cm) eluting with 10-50% EtOAc/hexanes to yield 4.4 g
(30 mmol, 98%) of the product. Analysis using a chiral GC
column showed the material was 88% ee. The product was
0.5 M Na2S2O3 (2×) and brine, dried (MgSO4), and concen-
trated under reduced pressure. The crude product was puri-
fied by silica gel flash chromatography (3 × 6 cm) eluting with
15% Et2O/pentane to yield 300 mg (1.39, mmol, 99%) of a
isolated as a white solid: mp ) 27-28° C; [R]23 ) -2.9 (c )
D
1.072, CHCl3); IR (neat) 3414, 2957, 2928, 2860, 1467, 1379,
colorless oil: [R]23 ) 13.6° (c ) 0.94, CHCl3); IR (neat) 2993,
D
1
1041 cm-1; H NMR (500 MHz, CDCl3) δ 3.76 (ddd, J ) 4.0,
2934, 2872, 2751, 1720, 1462, 1381 cm-1; 1H NMR (300 MHz,
CDCl3) δ 10.05 (d, J ) 4.0 Hz, 1 H), 4.14-4.03 (m, 3 H), 1.99-
1.97 (m, 1 H), 1.46 (s, 3 H), 1.37 (s, 3 H), 1.62-1.04 (m, 8 H),
0.830 (t, J ) 6.4 Hz, 3 H); 13C NMR (75 MHz, CDCl3, DEPT)
δ C, 99.3; CH, 201.4, 70.2, 49.7; CH2, 61.4, 33.2, 31.5, 24.9,
22.4; CH3, 29.4, 19.8, 13.9. Anal. Calcd for C12H22O2: C, 67.26;
H, 10.35. Found: C, 67.21; H, 10.15.
7.0, 11.6 Hz, 1 H), 3.57 (ddd, J ) 4.6, 7.0, 11.9 Hz, 1 H), 3.80
(dt, J ) 7.9, 4.3 Hz, 1 H), 2.99-2.93 (m, 2 H), 1.50-1.20 (m,
8 H), 0.810 (t, J ) 7.0 Hz, 3 H); 13C NMR (75 MHz, CDCl3,
DEPT) δ CH, 57.3, 57.0; CH2, 60.9, 31.6, 27.9, 26.3, 22.5; CH3,
13.9; HRMS (CI) calcd for C8H20NO2 162.1495, found 162.1493
(M + NH4). Anal. Calcd for C8H16O2: C, 66.63; H, 11.18.
Found: C, 66.59; H, 10.93.
(4a (S),5R)-2,2-Dim eth yl-4-p en tyl-R-(3-p r op en yl)-1,3-d i-
oxa n e-5-m eth a n ol (33). A 100 mL round bottom flask was
charged with aldehyde 32 (440 mg, 2.05 mmol) and 20 mL of
(3R,4R)-3-(Hyd r oxym eth yl)-1-n on en -4-ol. CuI (800 mg,
4.20 mmol) was suspended in 100 mL of Et2O under Ar at -10
°C. Vinylmagnesium bromide (42 mL of 1.0 M solution in
THF) was added via syringe. The solution was stirred for 10
min and then cooled to -25 °C. A solution of epoxide 30 (2.0
g, 13.9 mmol) in 5 mL of Et2O was added via cannula. After
the mixture was stirred under Ar at -25 °C for 9 h, the
reaction was quenched with 40 mL of saturated NH4Cl which
had been brought to pH 8 with NH4OH. The aqueous layer
was extracted (3 × 50 mL) with Et2O. The combined organic
layers were washed with brine, dried (MgSO4), and then
concentrated under reduced pressure. The crude product was
purified by silica gel flash chromatography (7 × 6 cm) eluting
with 50% EtOAc/hexanes. The mixture of diols was taken up
in 100 mL of 1:1 THF/water. A solution of NaIO4 (3.0 g, 14
mmol) in 20 mL of water was added. The reaction mixture
was stirred 3 h and then diluted with 100 mL of brine. The
aqueous phase was extracted (3 × 50 mL) with Et2O. The
combined organic layers were dried (MgSO4) and then con-
centrated under reduced pressure. The crude product was
purified by MPLC on SiO2 (30 × 4 cm) eluting with 50%
EtOAc/hexanes to give 1.19 g (6.92 mmol, 50%) of a colorless
oil: [R]24D ) 4.5° (c ) 1.05, CHCl3); IR (neat) 3371, 3076, 2956,
d
Et2O and cooled to -78 °C. A 1 M solution of Ipc2BAllyl in
pentane (2.5 mL, 2.5 mmol) was added to the reaction flask
via syringe. The reaction mixture was stirred for 21 h and
then warmed to 0 °C. The reaction was quenched by the slow
addition of 1 mL of 3 N NaOH and 2 mL of 30% H2O2, and
then the solution was heated to reflux for 1 h. The aqueous
layer was extracted (2 × 20 mL) with Et2O. The combined
organic layers were washed with saturated NaHCO3, water
(2×), and brine and dried (MgSO4). TLC showed incomplete
oxidation. The solution was concentrated to an oil and then
dissolved in 20 mL of THF followed by the slow addition of
1.5 mL of 3 N NaOH and 3 mL 30% H2O2. The solution was
refluxed for 30 min and then extracted (3 × 25 mL) with Et2O.
The combined organic layers were washed with saturated
NaHCO3, water (2×), and brine, dried (MgSO4), and concen-
trated to a colorless oil. The crude product was purified by
silica gel flash chromatography (5 × 6 cm) eluting with 10%
EtOAc in 1:1 CH2Cl2:hexanes to yield 440 mg (1.72, mmol,
84%) of a colorless oil: [R]26 ) 10.0° (c ) 1.01, CHCl3); IR
D
(neat) 3437, 3074, 2290, 2932, 2868, 1640, 1462, 1379 cm-1
;
1H NMR (500 MHz, CDCl3) δ 5.84 (dddd, J ) 5.5, 8.9, 10.1,
17.1 Hz, 1 H), 5.20-5.15 (m, 2 H), 4.07-3.94 (m, 4 H), 2.70
(ddddd, J ) 1.7, 1.7, 2.9, 5.6, 14.5 Hz, 1 H), 2.37 (dt, J ) 14.5,
9.6 Hz, 1 H), 1.98 (broad s, 1 H), 1.44 (s, 3 H), 1.34 (s, 3 H),
1.62-1.24 (m, 9 H), 0.878 (t, J ) 7.0 Hz, 3 H); 13C NMR (75
MHz, CDCl3, DEPT) δ C, 98.5; CH, 136.0, 71.3, 68.5, 10.8; CH2,
118.4, 61.1, 40.2, 33.0, 31.8, 25.6, 22.6; CH3, 29.7, 18.8, 14.1;
HRMS (CI) calcd for C15H29O3 257.2118, found 257.2112 (M +
H). Anal. Calcd for C15H28O3: C, 70.27; H, 11.01. Found: C,
70.16; H, 10.88.
2931, 2873, 1640, 1467 cm-1 1H NMR (500 MHz, CDCl3) δ
;
5.82 (ddd, J ) 8.9, 10.4, 17.4 Hz, 1 H), 5.21 (dd, J ) 1.8, 10.4
Hz, 1 H), 5.14 (ddd, J ) 0.9, 1.8, 17.4 Hz, 1 H), 3.80-3.68 (m,
3 H), 2.75 (t, J ) 4.9 Hz, 1 H), 2.58 (d, J ) 4.6 Hz, 1 H), 2.28
(dddd, J ) 3.4, 5.8, 8.9, 11.9 Hz, 1 H), 1.46-1.38 (m, 3 H),
1.31-1.22 (m, 5 H), 0.858 (t, J ) 6.7 Hz, 3 H); 13C NMR (75
MHz, CDCl3, DEPT) δ CH, 134.8, 72.7, 50.6; CH2, 118.7, 64.6,
34.8, 31.8, 25.5, 22.6; CH3, 14.0; HRMS (CI) calcd for C10H24
-
NO2 190.1808, found 190.1800 (M + NH4). Anal. Calcd for
C10H20O2: C, 69.72; H, 11.70. Found: C, 69.48; H, 11.54.
(3R,4R)-2,2-Dim et h yl-3-et h en yl-4-p en t yl-1,3-d ioxa n e
(31). (3R,4R)-3-(Hydroxymethyl)-1-nonen-4-ol (1.10 g, 6.39
mmol) was dissolved in 40 mL of acetone and 20 mL of DMP.
PPTS (100 mg, 0.4 mmol) was added. The reaction was tightly
capped under N2 and stirred for 16 h. The reaction was
quenched with 100 µL of Et3N, and then the solution was
concentrated under reduced pressure. The crude product was
purified by silica gel flash chromatography (6 × 6 cm) eluting
with 10% Et2O/pentane to yield 1.19 g (5.55 mmol, 87%) of a
[2R,4S,4(4R,5R)]- a n d [2S,4S,4(4R,5R)]-4-(2,2-Dim eth yl-
4-p en tyl-1,3-d ioxa n -5-yl)-2,4-d ih yd r oxy-2,4-O-(1-m eth yl-
eth ylid en e)bu ta n en itr ile (12). A 25 mL round bottom flask
was charged with homoallylic alcohol 33 (400 mg, 1.56 mmol),
10 mL of CH3CN, and bis(trimethylsilyl)acetamide (500 µL,
2.0 mmol). The reaction mixture was heated to reflux for 10
h. The solution was concentrated under reduced pressure to
an oil which was flashed though a plug of silica gel eluting
with 10% Et2O/pentane to yield 490 mg of a colorless oil. The
oil was dissolved in 10 mL of acetone and 3 mL of water in a
50 mL round bottom flask followed by the addition of NMO
(260 mg, 2.22 mmol) and a 2.5% solution of OsO4 in tert-butyl
alcohol (1.10 mL, 0.09 mmol). The reaction mixture was
stirred for 1 h followed by the addition of a solution of NaIO4
(650 mg, 3.0 mmol) in 7 mL of water. The mixture was stirred
for 30 min during which a white precipitate formed. The
solution was extracted (3 × 25 mL) with Et2O. The organic
phase was washed with 0.5 M Na2S2O3 (2×) and brine, dried
(MgSO4), and concentrated under reduced pressure to an oil
which was flashed though a plug of silica gel eluting with 15%
Et2O/pentane to yield 413 mg of a colorless oil. The oil was
placed in a 25 mL round bottom flask, flushed with argon, and
cooled to 0 °C. TMSCN (230 µL, 1.90 mmol) and 1 mg of KCN/
18-crown-6 complex were added. The ice bath was removed
and the reaction mixture stirred for 1.5 h followed by the
addition of a solution of 20 mg of CSA in 15 mL of 2:1 acetone/
DMP. The reaction mixture was stirred for 16 h, and then
the reaction was quenched with 20 µL of Et3N. The solution
was concentrated under reduced pressure to an oil which was
flashed though a plug of silica gel eluting with 10% EtOAc/
hexanes to yield 373 mg (1.15 mmol, 74%) of a colorless oil:
colorless oil: [R]22 ) 6.4° (c ) 0.98, CHCl3); IR (neat) 3073,
D
2992, 2956, 2936, 2860, 1641, 1461 cm-1; 1H NMR (500 MHz,
CDCl3) δ 6.21 (dt, J ) 17.4, 10.1 Hz, 1 H), 5.14-5.05 (m, 2 H),
4.12 (dd, J ) 3.1, 11.3 Hz, 1 H), 3.89 (m, 1 H), 3.69 (dd, J )
1.5, 11.3 Hz, 1 H), 1.94 (m, 1 H), 1.44 (s, 3 H), 1.39 (s, 3 H),
1.41-1.19 (m, 8 H), 0.857 (t, J ) 7.0 Hz, 3 H); 13C NMR (75
MHz, CDCl3, DEPT) δ C, 98.7; CH, 136.0, 71.1, 43.1; CH2,
116.6, 65.9, 33.4, 31.7, 24.6, 22.6; CH3, 29.6, 19.1, 14.0; HRMS
(CI) calcd for C13H25O2 213.1856, found 213.1853 (M + H).
Anal. Calcd for C13H24O2: C, 73.54; H, 11.39. Found: C,
73.31; H, 11.56.
(4R,5R)-2,2-Dim et h yl-5-et h en yl-4-p en t yl-1,3-d ioxa n e
(32). Acetonide 31 (300 mg, 1.4 mmol) was dissolved in 10
mL of acetone and 3 mL of water in a 50 mL round bottom
flask followed by the addition of NMO (245 mg, 2.1 mmol) and
a 2.5% solution of OsO4 in tert-butyl alcohol (900 µL, 0.07
mmol). The reaction mixture was stirred for 6 h followed by
the addition of a solution of NaIO4 (900 mg, 4.2 mmol) in 7
mL of water. The mixture was stirred for 2 h during which
time a white precipitate formed. The solution was extracted
(3 × 25 mL) with Et2O. The organic phase was washed with

Filipin III
J . Org. Chem., Vol. 61, No. 13, 1996 4231
1
IR (neat) 2991, 2936, 2871, 2252, 1462, 1381 cm-1; H NMR
(500 MHz, CDCl3) δ 4.76 (dd, J ) 2.7, 11.7 Hz, 1 H), 4.26 (ddd,
J ) 2.4, 5.8, 11.7 Hz, 1 H), 4.03-3.90 (m, 3 H), 2.19-2.03 (m,
2 H), 1.69 (s, 1 H), 1.46 (s, 3 H), 1.42 (s, 3 H), 1.39 (s, 3 H),
1.33 (s, 3 H), 1.54-1.21 (m, 8 H), 0.873 (t, J ) 7.0 Hz, 3 H);
13C NMR (75 MHz, CDCl3, DEPT) δ C, 118.0, 99.9, 98.7; CH,
71.3, 65.8, 59.6, 40.9; CH2, 60.7, 33.1, 32.9, 31.7, 25.3, 22.6;
CH3, 30.1 (2), 19.2, 18.8, 14.1; HRMS (CI) calcd for C18H35N2O4
[1(4R,5R),1S,3R,5S,7R,9S,11R,13S,14S]-1-(2,2-Dim eth yl-
4-p en tyl-1,3-d ioxa n -5-yl)-1,3:5,7:9,11:13,14-tetr a k is-O-(1-
m eth yleth yliden e)pen tadecan -1,3,5,7,9,11,13,14-octol (10).
A 25 mL round bottom flask equipped with a cold finger was
cooled to -78 °C and charged with 5 mL of NH3 and Li (6 mg,
0.86 mmol). A solution of pentaacetonide trinitrile 34 (30 mg,
0.039) in 1 mL of THF was transferred to the reaction flask
via cannula. The reaction mixture was stirred for 1 h, and
then the reaction was quenched with 200 mg of NH4Cl. The
mixture was allowed to warm to room temperature and the
NH3 allowed to evaporate. Water was then added and the
solution extracted (3 × 10 mL) with CH2Cl2. The combined
organic phases were washed with brine, dried (Na2SO4), and
concentrated under reduced pressure. The crude product was
purified by silica gel flash chromatography (2 × 4 cm) eluting
with 30% EtOAc/hexanes to yield 21 mg (0.031 mmol, 79%) of
343.2597, found 343.2596 (M + NH4). Anal. Calcd for C18H31
-
NO4: C, 66.43; H, 9.60. Found: C, 66.59; H, 9.46.
[1(4R,5R),1S,3S,5R,7R,9S,11S,13S,14S]-1-(2,2-Dim eth yl-
4-p en tyl-1,3-d ioxa n -5-yl)-3,7,11-tr icya n o-1,3:5,7:9,11:13,-
14-tetr a k is-O-(1-m eth yleth ylid en e)p en ta d eca n -1,3,5,7,9,-
11,13,14-octol (34). A solution of cyanohydrin acetonide 12
(87 mg, 0.267 mmol) in 200 µL of THF was added via syringe
to 0.324 mmol of LiNEt2 in 200 µL of THF under Ar at -78
°C. The reaction mixture was stirred for 1 h, and then the
solution was warmed to -63 °C and stirred for another hour.
A solution of iodide 27 (50 mg, 0.089 mmol) in 200 µL of THF
was added to the reaction flask via syringe. The solution was
warmed to -25 °C with a MeOH/ice bath and allowed to warm
slowly to 10 °C over 14 h. The reaction was quenched with
20 mL of saturated NH4Cl. The aqueous layer was extracted
(3 × 25 mL) with CH2Cl2. The combined organic layers were
washed with brine, dried (Na2SO4), and then concentrated
under reduced pressure. The crude product was purified by
silica gel flash chromatography (3 × 5 cm) eluting with 20%
tert-butyl methyl ether/hexanes to yield 35 mg of recovered
cyanohydrin acetonide 12 and 45 mg (0.059 mmol, 66%) of a
a colorless oil: [R]26 ) -19.2° (c ) 1.05, CHCl3); IR (neat)
D
2989, 2938, 2868, 1460, 1436, 1379 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.26 (ddd, J ) 2.0, 6.0, 11.8 Hz, 1 H), 4.06-3.95 (m,
7 H), 3.89 (dd, J ) 3.0, 12.1 Hz, 1 H), 3.68-3.66 (m, 2 H),
1.84-1.77 (m, 2 H), 1.73-1.10 (m, 19 H), 1.43 (s, 3 H), 1.42 (s,
6 H), 1.41 (s, 3 H), 1.37 (s, 3 H), 1.36 (s, 6 H), 1.35 (s, 3 H),
1.33 (s, 3 H), 1.32 (s, 3 H), 1.24 (d, J ) 5.1 Hz, 3 H), 0.878 (t,
J ) 6.7 Hz, 3 H); 13C NMR (75 MHz, CDCl3, DEPT) δ C, 107.7,
98.5, 98.4, 98.3, 98.2; CH, 79.0, 77.1, 71.6, 66.7, 66.6, 65.7, 65.3,
65.2, 65.1, 41.1; CH2, 61.0, 42.7(2), 39.6, 37.4, 36.5, 34.8, 33.1,
31.9, 25.3, 22.6; CH3, 30.3, 30.25, 30.23, 29.7, 27.3(2), 20.0,
colorless oil: [R]26 ) 23.2° (c ) 1.80, CHCl3); IR (neat) 2991,
D
19.9, 19.8, 18.9, 17.2, 14.1; HRMS (FAB) calcd for C38H69O10
-
2933, 2872, 2253, 1459, 1433, 1381 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.59-4.45 (m, 3 H), 4.04-3.99 (m, 2 H), 3.93 (dd, J
) 3.1, 12.3 Hz, 1 H), 3.81 (dt, J ) 3.1, 8.2 Hz, 1 H), 3.73 (dq,
J ) 5.8, 8.2 Hz, 1 H), 2.44 (dd, J ) 12.0, 14.0 Hz, 1 H), 2.10-
1.90 (m, 8 H), 1.77 (dd, J ) 1.8, 13.4 Hz, 1 H), 1.72 (s, 6 H),
1.70 (s, 3 H), 1.61 (dd, J ) 11.8, 13.9 Hz, 1 H), 1.59-1.28 (m,
10 H), 1.42 (s, 3 H), 1.40 (s, 3 H), 1.39 (s, 3 H), 1.37 (s, 6 H),
1.35 (s, 3 H), 1.32 (s, 3 H), 1.28 (d, J ) 5.8 Hz, 3 H), 0.883 (t,
J ) 6.9 Hz, 3 H); 13C NMR (75 MHz, CDCl3, DEPT) δ C, 122.1,
121.7, 121.0, 108.7, 101.5, 101.1(2), 98.5, 69.5, 68.3, 68.1; CH,
77.8, 76.9, 70.4, 64.7, 61.9(2), 39.8; CH2, 59.8, 46.4, 46.3, 45.2,
40.0, 39.8, 38.1, 36.2, 31.7, 25.3, 22.6; CH3, 31.0, 30.8, 30.7,
29.9, 27.3, 27.1, 21.6, 21.4, 21.2, 18.6, 17.0, 14.0; HRMS (FAB)
calcd for C41H66N3O10 760.4751, found 760.4760 (M + H). Anal.
Calcd for C41H65N3O10: C, 64.80; H, 8.62. Found: C, 65.00;
H, 8.82.
Na 707.4698, found 707.4710 (M + Na). Anal. Calcd for
C38H68O10: C, 67.01; H, 9.81. Found: C, 66.64; H, 10.01.
Ack n ow led gm en t. Support has been provided by
the National Institutes of Health, the Camille and
Henry Dreyfus Foundation Teacher-Scholar Award,
and the Alfred P. Sloan Foundation. Additional support
was provided by American Cyanamid, Bristol-Myers
Squibb, Pfizer, and Zeneca. The generous donation of
filipin complex by the Upjohn Co. is gratefully
acknowledged.
J O960218X