Synthesis of the bicyclic core of the nucleoside antibiotic octosyl acid A

Article abstract of DOI:10.1021/jo052025s

The bicyclic core of octosyl acid A has been prepared using a diastereoselective acetylide addition and 6-endo selenoetherification as key steps. A detailed study of the selenoetherification reaction and difficulties encountered in the conversion of a phe

Full text of DOI:10.1021/jo052025s

Synthesis of the Bicyclic Core of the Nucleoside Antibiotic Octosyl
Acid A
Jesse D. More*,† and Nathaniel S. Finney*,‡
Department of Chemistry and Biochemistry, UniVersity of California, San Diego, 9500 Gilman DriVe,
La Jolla, California, 92093
finney@unizh.ch
ReceiVed September 27, 2005
The bicyclic core of octosyl acid A has been prepared using a diastereoselective acetylide addition and
6-endo selenoetherification as key steps. A detailed study of the selenoetherification reaction and difficulties
encountered in the conversion of a phenyl group to a carboxylic acid will be discussed.
CHART 1
Introduction
Certain bacterial species of the genus Streptomyces produce
structurally and biologically interesting higher-carbon sugars.1
Salient structural features of these metabolites include stereo-
chemically rich cis- or trans-fused furopyran ring systems and
a natural or modified pyrimidine base. This group of natural
products (see Chart 1) includes octosyl acids A, 1, and C, 2;2
the C-glycoside malayamycin A, 3;3 and the more complex
ezomycins.4 Although malayamycin A and the ezomycins
display antifungal activity, the octosyl acids do not show any
biological activity of their own. An adenine analogue of octosyl
acid A, however, was shown to be a competitive inhibitor of
cyclic nucleotide phosphodiesterases from various animal
tissues.2b
furanose scaffold. In a pioneering effort, Danishefsky and
Hungate used an intramolecular O-alkylation of a stannylene
acetal to construct the six-membered ring of 1.5a Shortly
thereafter, Hanessian reported an approach to 1 that relied on a
6-exo intramolecular oxymercuration as the key ring-forming
step.5b Several other partial synthetic approaches to 1 have been
reported using either alkylation5c,6a,b or radical6c,d annulation
strategies.
During our synthetic studies on the polyoxins,7 we found that
Carreira’s zinc acetylide addition8 allows for highly diastereo-
selective homologation of C5 of ribose (see Scheme 1). This
matched diastereoselective operation provides most of the carbon
skeleton of 1. Annulation of the pyran ring (via a 6-endo
pathway) onto the furanose scaffold followed by N-glycosylation
Despite modest medicinal potential, octosyl acid A has
received attention from the synthetic community culminating
in two syntheses.5,6 Most of the synthetic efforts directed toward
1 have involved annulating the pyran ring onto an existing
† Present address: Department of Chemistry, Loyola College, 4501 N. Charles
Street, Baltimore, Maryland, 21210.
‡ Present address: Organisch-Chemisches Institut, Universita¨t Zu¨rich, Win-
terthurerstrasse 190, 8057 Zu¨rich.
(1) Knapp, S. Chem. ReV. 1995, 95, 1859-1876.
(2) (a) Isono, K.; Crain, P. F.; McCloskey, J. A. J. Am. Chem. Soc. 1975,
97, 943-945. (b) Sato, T.; Hirasawa, K.; Uzawa, J.; Inaba, T.; Isono, K.
Tetrahedron Lett. 1976, 1687-1690.
(3) Hanessian, S.; Marcotte, S.; Machaalani, R.; Huang, G. Org. Lett.
2003, 5, 4277-4280.
(6) For synthetic studies, see: (a) Anzai, K.; Saita, T. Bull. Chem. Soc.
Jpn. 1977, 50, 169-174. (b) Kim, K. S.; Szarek, W. A. Carbohydr. Res.
1982, 100, 169-176. (c) Araki, Y.; Endo, T.; Arai, Y.; Tanjo, M.; Ishido,
Y. Tetrahedron Lett. 1989, 30, 2829-2832. (d) Malolanarasimhan, K. Ph.D.
Dissertation, Rutgers University, New Brunswick, NJ, 2001.
(7) More, J. D.; Finney, N. S. Synlett 2003, 1307-1310.
(4) Sakata, K.; Sakurai, A.; Tamura, S. Agric. Biol. Chem. 1973, 37,
697-699.
(5) (a) Danishefsky, S. J.; Hungate, R. J. Am. Chem. Soc. 1986, 108,
2486-2487. (b) Hanessian, S.; Kloss, J.; Sugawara, T. J. Am. Chem. Soc.
1986, 108, 2758-2759. (c) Kozaki, S.; Sakanaka, O.; Yasuda, T.; Shimizu,
T.; Ogawa, S.; Suami, T. J. Org. Chem. 1988, 53, 281-286.
(8) Frantz, D. E.; Fassler, R.; Tomooka, C. S.; Carreira, E. M. Acc. Chem.
Res. 2000, 33, 373-381.
10.1021/jo052025s CCC: $33.50 © 2006 American Chemical Society
Published on Web 02/23/2006
2236
J. Org. Chem. 2006, 71, 2236-2241

Synthesis of the Bicyclic Core of Octosyl Acid A
SCHEME 1
SCHEME 4
SCHEME 2
would provide a concise route to 1. Several alternatives for the
annulation were considered; electrophile-mediated etherification
(cf. Scheme 2, 6f7) was deemed most promising. Our efforts
toward the synthesis of octosyl acid A, which have resulted in
a stereocontrolled preparation of the bicyclic core of the natural
product, are described in detail below.9
SCHEME 5
Results and Discussion
Our synthetic plan revolved around electrophile-mediated
etherification of a suitably substituted olefin 6 (see Scheme 2).10
A similar approach to octosyl nucleosides has been reported11
but suffers from the need for correction of the configuration at
C5′ (octosyl acid A numbering used throughout) and manipula-
tion of the pyrimidine ring to arrive at the natural product. Our
approach generates the proper configuration at C5′ and allows
for incorporation of the 5-carboxy uracil directly via glycosy-
lation. We initially decided to focus on 8 as a specific
embodiment of 6, as it contains the C8′ carboxylate moiety in
conveniently protected form.
formation, a second hydropalladation, and palladium(II) π-allyl
capture by an acetate nucleophile. Our successful adaptation of
this reaction is shown in Scheme 4. Addition of the zinc
acetylide derived from propargyl acetate to aldehyde 47 followed
by silylation afforded 9 as a single diastereomer. In line with
the precedent mentioned above, treatment of this compound with
Pd(PPh3)4 (generated in situ) and acetic acid in boiling toluene
gave gem-diacetate 10 in 55-74% yield after silica gel
chromatography. This formal internal redox reaction was
efficient on scales up to 10 g. A three-step sequence of
methanolysis, Pinnick oxidation,14 and esterification gave the
unsaturated ester 12 (via aldehyde 11) in excellent overall yield.
Removal of the acetonide from 12 in acidic methanol (not
shown) was accompanied by loss of the silyl group, necessitating
a protecting group exchange (12f14, Scheme 5). The acetonide
in 14 could now be cleanly removed to give a separable mixture
of anomers 15 in 81% combined yield. Despite considerable
experimentation, we were unable to generate the fused furopyran
system from 15 (Scheme 6). We hypothesized that replacing
the methoxycarbonyl substituent with a more electron-rich group
(Y in structure 8) would render the olefin more nucleophilic
and thus more reactive with an added electrophile.
Our synthesis of 8 is based on a reaction discovered by Trost12
(and later used by Carreira) which converts propargyl acetates
into γ-hydroxy R-â-unsaturated aldehydes (Scheme 3).13 The
SCHEME 3
precise mechanism is not known but likely involves regio-
selective hydropalladation of the alkyne, reductive allene
(9) More, J. D. Ph.D. Dissertation, University of California, San Diego,
La Jolla, CA, March, 2004.
(10) For a review of electrophile-induced cyclizations, see: Frederickson,
M.; Grigg, R. Org. Prep. Proced. Int. 1997, 29, 33-62.
(11) Haraguchi, K.; Hosoe, M.; Tanaka, H.; Tsuruoka, S.; Kanmuri, K.;
Miyasaka, T. Tetrahedron Lett. 1998, 39, 5517-5520.
(12) Trost, B. M.; Brieden, W.; Baringhaus, K. H. Angew. Chem., Int.
Ed. Engl. 1992, 31, 1335-1336.
The use of a phenyl group as a carboxylic acid surrogate
(see 17) was investigated next. The electron-donating nature of
(13) El-Sayed, E.; Anand, N. K.; Carreira, E. M. Org. Lett. 2001, 3,
3017-3020.
(14) Bal, B. S.; Childers, W. E., Jr.; Pinnick, H. W. Tetrahedron 1981,
37, 2091-2096.
J. Org. Chem, Vol. 71, No. 6, 2006 2237

More and Finney
SCHEME 6
SCHEME 8
the phenyl group (relative to an ester) should increase the
nucleophilicity of the olefin thereby facilitating the formation
of an “onium” ion intermediate such as A. In addition, the
aromatic ring should stabilize, via resonance, an adjacent
developing positive charge (cf. A, Y ) phenyl) and facilitate
cyclization. Although oxidation of a phenyl ring to a carboxylic
acid is a difficult transformation, it has been achieved on several
complex substrates using the Sharpless modification of the
classic Djerassi oxidant, RuO4.15,16 This approach seemed
especially attractive as we had prepared substrates such as 17
during our polyoxin synthetic studies.7
SCHEME 9
All attempts to remove the acetonide from known olefin 189
resulted in SN1 substitution by methanol (Scheme 7). Given this
SCHEME 10
SCHEME 7
(along with unreacted starting material) to which we assigned
the structure 26 (Scheme 9). Spectral data supporting this
conclusion included the high-resolution mass spectrum (M +
Na peak at 487.0642) and the 1H NMR spectrum, which showed
no peaks in the olefinic region and 10 aromatic protons. The J
value of 11 Hz between the doublet of doublets at 3.31 ppm
result, deprotection of the acetonide was performed at the
propargylic ester stage, thereby preventing SN1-type chemistry
by destabilizing a putative carbocation. Starting from known
alcohol 20,7 benzoate protection and acetonide removal gave
22 (Scheme 8). The resulting mixture of anomers was separated
and carried forward separately for ease of characterization. After
protection of the major anomer as the bis(silyl) ether, treatment
with lithium aluminum hydride in warm THF smoothly removed
the benzoate and reduced the alkyne in one pot. Acylation and
fluoride treatment provided the cyclization substrate 25 in
excellent overall yield.
With diol 25 in hand, conditions for electrophile-induced
cyclization were explored. N-Iodosuccinimide gave no reaction,
and phenylselenyl chloride (PhSeCl)17 gave mainly decomposi-
tion of the starting material (not shown). Further experimentation
revealed that, in the presence of solid sodium bicarbonate, olefin
25 reacted with PhSeCl to give a new compound in 22% yield
(assigned to Hb) and the doublet at 4.87 ppm (assigned to Ha)
is indicative of a trans-diaxial relationship. Reaction with acetic
anhydride (Scheme 10) gave diacetate 27, with nOe difference
measurements (see arrows on structure 27) which lent further
support to the stereochemical assignment of selenide 26.
To optimize the conversion and yield of the selenoetherifi-
cation, we systematically varied the reaction conditions (Table
1). Addition of PhSeCl in one portion to a stirred suspension
of the olefin 25 and sodium bicarbonate in methylene chloride
at 0 °C gave a 22% yield of 26 (entry 1). Slow addition of
PhSeCl as a solution in methylene chloride under otherwise
identical conditions gave a slightly higher yield (entry 2). The
addition of molecular sieves (entry 3) increased the yield
slightly, but the product was contaminated with some byprod-
ucts, even after chromatography. Finally, it was found that slow
addition of PhSeCl at 0 °C, followed by warming to room
temperature, gave a reproducible 41% yield of the fused
oxacycle 26 (entry 4). Although the inclusion of solid sodium
bicarbonate was crucial to suppress acid-mediated decomposi-
tion, soluble bases (triethylamine with or without silver tri-
fluoroacetate or 2,6-lutidine) completely inhibited the reaction
(entries 5-7). An interesting solvent effect was observed. Of
several solvents surveyed, only methylene chloride gave any
appreciable level of conversion (entries 8-11). In a study of
(15) (a) Carlsen, P. H. J.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J.
Org. Chem. 1981, 46, 3936-3938. (b) Djerassi, C.; Engle, R. R. J. Am.
Chem. Soc. 1953, 75, 3838-3840.
(16) (a) Nicolaou, K. C.; Roecker, A. J.; Monenschein, H.; Guntupalli,
P.; Follman, M. Angew. Chem., Int. Ed. Engl. 2003, 42, 3637-3642. (b)
Boger, D. L.; Lee, R. J.; Bounandm, P.-Y.; Meier, P. J. Org. Chem. 2000,
65, 6770-6772.
(17) (a) Nicolaou, K. C.; Magolda, R. L.; Sipio, W. J.; Barnette, W. E.;
Lysenko, Z.; Joullie, M. M. J. Am. Chem. Soc. 1980, 102, 3784-3793. (b)
Nicolaou, K. C.; Petasis, N. A. Selenium in Natural Products Synthesis;
CIS, Inc.: Philadelphia, 1984.
2238 J. Org. Chem., Vol. 71, No. 6, 2006

Synthesis of the Bicyclic Core of Octosyl Acid A
TABLE 1. Optimization of Selenoetherification
SCHEME 12
yield
entry solvent
reaction conditions
CH2Cl2 NaHCO3 (6 equiv), 0 °C
CH2Cl2 NaHCO3 (10 equiv), 0 °C
(%)a,b
1
2
22c
35 (63)d
40
3
4
5
6
CH2Cl2 NaHCO3 (10 equiv), 0 °C, 4A MS
CH2Cl2 NaHCO3 (10 equiv), 0 °C to room temp
CH2Cl2 NEt3 (5 equiv), 0 °C
41 (58)d
NRc
CH2Cl2 NEt3 (5 equiv), Ag(O2CCF3) (2 equiv), 0 °C NRc
7
8
CH2Cl2 2,6-lutidine (4 equiv), 0 °C
THF NaHCO3 (10 equiv), room temp
NRc
NR
conditions recently described by Vasconcellos for oxidation of
aromatic rings in the presence of cyclic benzylic ethers and,
importantly, in the absence of bulk water.20 Under the described
conditions (periodic acid, RuCl3, CH3CN), we observed acid-
mediated opening of the ketal to give 30 (based on 1H NMR of
the crude reaction mixture). Addition of molecular sieves under
otherwise identical conditions returned the starting material. This
nonaqueous oxidation protocol holds promise, however, as
benzylic oxidation was not observed.
9
10
11
CH3CN NaHCO3 (10 equiv), room temp
PhCH3 NaHCO3 (10 equiv), room temp
MeOH NaHCO3 (10 equiv), 0°C to room temp
NR
NR
complex
mixture
a NR ) no reaction. b Slow addition (dropwise) of PhSeCl unless
otherwise noted. c PhSeCl added in one portion. d Yield based on recovered
starting material.
SCHEME 11
Conclusions
We have described a synthesis of the trans-fused furopyran
ring system of octosyl acid A 1 and related natural products.
From D-ribose, we used a short sequence of highly diastereo-
selective operations (Carreira acetylide addition, 6-endo sele-
noetherification) to access the desired ring system. Our synthetic
route was ultimately derailed by the inability to convert a phenyl
group into a carboxylate. A successful synthetic route will
require modification of the Carreira protocol to include an alkyne
substituent more readily converted to a carboxylic acid, such
as a more electron-rich aromatic ring.
selenium-mediated tetrahydrofuran formation,18 Lipshutz and
co-workers also observed a narrow solvent scope for such
reactions. However, their optimal solvent was acetonitrile, and
no cyclization was observed in any other solvent, including
methylene chloride. The origin of our solvent effect is still
unclear.
Experimental Section
Benzoate 21. Propargyl alcohol 207 (5.53 g, 18.20 mmol, 1
equiv) and 4-DMAP (0.111 g, 0.910 mmol, 0.050 equiv) were dried
by coevaporation with benzene (2×) and then dissolved in CH2Cl2
(100 mL) and cooled to 0 °C. Triethylamine (3.83 mL, 27.30 mmol,
1.50 equiv) was added, followed by benzoyl chloride (2.20 mL,
19.10 mmol, 1.05 equiv). The reaction was removed from the ice
bath after 15 min and, after a total reaction time of 2 h, was poured
onto ice and saturated aqueous NaHCO3, stirred until the ice melted,
and then extracted with CH2Cl2 (3×). It was washed with 1 M
HCl, saturated aqueous NaHCO3, and brine and was dried (sodium
sulfate) and concentrated to yield a thick yellow oil. 21 (6.80 g,
92%) thus obtained was of sufficient purity for subsequent reactions.
TLC Rf (30% EtOAc/hexanes): 0.40. 1H NMR (400 MHz, CDCl3)
δ: 8.11 (m, 2 H), 7.59 (m, 1 H), 7.47 (m, 4 H), 7.33-7.27 (m, 3
H), 5.93 (d, 1 H, J ) 7.6 Hz), 5.05 (s, 1 H), 4.90 (d, 1 H, J ) 6.0
Hz), 4.67 (d, 1 H, J ) 6.0 Hz), 4.60 (d, 1 H, J ) 7.6 Hz), 3.36 (s,
3 H), 1.54 (s, 3 H), 1.36 (s, 3 H). 13C NMR (100 MHz, CDCl3) δ:
165.0, 133.2, 131.9, 129.8, 129.4, 128.7, 128.3, 128.1, 121.9, 112.6,
110.1, 87.3, 86.6, 85.3, 84.2, 81.6, 65.1, 55.6, 26.6, 25.2. FTIR
(thin film) cm-1: 2966 (w), 2950 (w), 1729 (s), 1456 (w), 1384
(m), 1264 (s), 1116 (s), 10956 (s), 1072 (m). HRMS (MALDI-
FTMS) m/z: calcd for C24 H24 O9 (M+ + Na), 431.1465; found,
431.1460.
Having obtained 27, it was necessary to remove the sele-
nophenyl group, convert the C8′ phenyl group to a carboxylic
acid, and install the 5-carboxy uracil ring to reach 1. The first
of these tasks was accomplished with excess tris(trimethylsilyl)-
silane19 and catalytic AIBN in reluxing toluene to yield 28 in
high yield (Scheme 11). Thorough degassing of the reaction
mixture was crucial. Oxidation of a phenyl ring to a carboxylic
acid has, as mentioned above, been carried out on fairly complex
molecules using the Sharpless protocol.15,16 Unfortunately, in
this case, all attempts to degrade the phenyl ring failed, despite
variation of the ruthenium source (RuCl3, RuO2), the oxidant
(NaIO4, KIO4, Bu4NIO4), and the solvent (EtOAc, CCl4)
(Scheme 11). Added base (NaHCO3) had no beneficial effect
on the reaction.
1
Crude H NMR spectra indicate a mixture of products but
are consistent with benzylic oxidation at C7′ to give a hemiacetal
(see Scheme 12). Acetal opening under the aqueous reaction
conditions gives an electron-poor aromatic ring (see 29) which
is immune to further oxidation. We also looked briefly at the
Methyl Glycosides 22. Benzoate 21 (6.76 g, 16.60 mmol, 1
equiv) was dissolved in MeOH (70 mL) and THF (20 mL). BioRad
(18) Lipshutz, B. H.; Barton, J. C. J. Am. Chem. Soc. 1992, 114, 1084-
1086.
(19) (a) Ballestri, M.; Chatgilialoglu, C.; Clark, K. B.; Griller, D.; Giese,
B.; Kopping, B. J. Org. Chem. 1991, 56, 678-683. (b) For a review, see:
Studer, A.; Amrein, S. Synthesis 2002, 835-849.
(20) Miranda, L. S. M.; Vasconcellos, M. L. A. A. Synthesis 2004, 1767-
1770.
J. Org. Chem, Vol. 71, No. 6, 2006 2239

More and Finney
50 W H+ resin (sulfonic-acid-type, 13.52 g, 2.00 equiv w/w) was
added, and the reaction was stirred at 70 °C for 19 h. At this time,
the reaction was cooled to room temperature. The resin was
removed by filtering through Celite, and the filtrate was concen-
trated. Chromatography (silica, 50% EtOAc/hexanes, 8 in. × 70
mm column) gave the pure â-anomer (3.23 g, 53%) as a white
solid (mp 107-109 °C), as well as a mixture of anomers (1.43 g,
76% overall) as an oil. Data for the â-anomer is given. TLC Rf
(50% EtOAc/hexanes): 0.21. 1H NMR (400 MHz, CDCl3) δ: 8.13
(m, 2 H), 7.59 (m, 1 H), 7.46 (m, 4H), 7.32 (m, 3 H), 6.12 (d, 1 H,
4.0 Hz), 4.90 (s, 1 H), 4.78 (t, 1 H, J ) 3.6 Hz), 4.40 (dd, 1 H, J
) 6.8, 4.0 Hz), 4.16 (d, 1 H, J ) 4.8 Hz), 3.21 (s, 3 H), 2.72 (br
s, 1 H), 2.58 (br s, 1 H). 13C NMR (100 MHz, CDCl3) δ: 165.2,
133.2, 131.8, 129.8, 129.4, 128.8, 128.2, 121.5, 108.0, 87.2, 83.6,
83.0, 71.7, 64.6, 55.4. FTIR (thin film) cm-1: 3442 (m), 2935 (w),
1724 (s), 1494 (w), 1452 (w), 1270 (s), 1110 (s), 1098 (s), 1028
(s, 3 H). 13C NMR (100 MHz, CDCl3) δ: 169.8, 136.2, 135.6,
128.4, 127.9, 126.5, 122.6, 108.4, 83.4, 76.2, 74.4, 72.4, 55.5, 25.94,
25.92, 21.5, 18.2, 18.1, -3.9, -4.3, -4.4, -4.6. FTIR (thin film)
cm-1: 2935 (s), 2862 (m), 1745 (s), 1477 (m), 1378 (m), 1255 (s),
1238 (s), 1162 (s), 1054 (m), 842 (m). HRMS (MALDI-FT MS)
m/z: calcd for C28 H48 O6 Si2 (M+ + Na), 559.2881; found,
559.2882.
Diol 25. The bis-TBS ether 24 (3.60 g, 6.70 mmol, 1 equiv)
was dissolved in THF (40 mL) and cooled to 0 °C, and TBAF (1.0
M in THF, 16.80 mL, 16.80 mmol, 2.50 equiv) was added dropwise
via pipet. The solution was stirred at 0 °C for 1 h and then warmed
to room temperature and quenched with saturated aqueous NaHCO3
after a 3 h reaction time. This mixture was extracted with EtOAc
(3×), washed with brine, dried (sodium sulfate), and concentrated.
Chromatography (silica, 60% EtOAc/hexanes, 40 mm × 7 in.) gave
25 (1.91 g, 92%) as a white solid (mp 91-94 °C). TLC Rf (60%
(m). HRMS (MALDI-FTMS) m/z: calcd for C21 H20 O6 (M+
+
1
EtOAc/hexanes): 0.18. H NMR (400 MHz, CDCl3) δ: 7.40 (m,
Na), 391.1152; found, 391.1155.
2 H), 7.32 (m, 3 H), 6.72 (d, 1 H, J ) 16.0 Hz), 6.24 (dd, 1 H, J
) 16.0, 7.2 Hz), 5.64 (ddd, 1 H, J ) 7.6, 4.4, 0.8 Hz), 4.86 (s, 1
H), 4.38 (dd, 1 H, J ) 6.4, 5.2 Hz), 4.15 (dd, 1 H, J ) 6.8, 4.4
Hz), 4.04 (d, 1 H, J ) 4.8 Hz), 3.38 (s, 3 H), 2.14 (s, 3 H). 13C
NMR (100 MHz, CDCl3) δ: 170.1, 135.9, 134.6, 128.4, 128.0,
126.6, 123.0, 108.3, 84.0, 75.1, 74.8, 71.8, 55.5, 21.4. FTIR (thin
film) cm-1: 3407 (m), 2935 (w), 1739 (s), 1372 (m), 1241 (s),
1113 (m), 1039 (m). HRMS (MALDI-FT MS) m/z: calcd for C16
H20 O6 (M+ + Na), 331.1152; found, 331.1152.
Bis-TBS Ether 23. Diol 22â (3.20 g, 8.70 mmol, 1 equiv) was
dried by coevaporation with benzene (2×) and then dissolved in
CH2Cl2 (75 mL) and cooled to 0 °C. 2,6-Lutidine (5.06 mL, 43.40
mmol, 5.00 equiv) was added followed by the dropwise addition
of TBSOTf (4.39 mL, 19.10 mmol, 2.20 equiv). After stirring for
3.5 h at 0 °C, the reaction was quenched with H2O and then warmed
to room temperature. The organic layer was removed, and the
aqueous layer was extracted with CH2Cl2 (2×), washed with
saturated aqueous NaHCO3, 1 M HCl, and brine, and dried (sodium
sulfate) and concentrated. Chromatography (silica, 10% EtOAc/
hexanes, 6 in. × 40 mm column) gave 23 (4.99 g, 96%) as a clear
Bicyclic Ether 26. Diol 25 (0.097 g, 0.315 mmol, 1 equiv) was
dissolved in CH2Cl2 (3 mL), and powdered NaHCO3 (0.264 g, 3.14
mmol, 10.0 equiv) was added. The resulting suspension was stirred
vigorously at room temperature. PhSeCl (0.066 g, 0.346 mmol, 1.10
equiv) was added dropwise in CH2Cl2 (1 mL + 0.50 mL rinse)
over 20 min. The reaction was stirred for 19 h and then filtered
through Celite and concentrated. The resulting yellow oil was
purified by chromatography (silica, 40% EtOAc/hexanes to 70%
EtOAc/hexanes, 5 in. × 25 mm column) to yield 26 (0.060 g, 41%)
as an oil plus recovered starting material (0.028 g). TLC Rf (50%
1
oil. TLC Rf (10% EtOAc in hexanes): 0.42. H NMR (400 MHz,
CDCl3) δ: 8.15 (m, 2 H), 7.58 (m, 1 H), 7.45 (m, 4 H), 7.31 (m,
3 H), 6.09 (d, 1 H, J ) 2.4 Hz), 4.78 (s, 1 H), 4.51 (dd, 1 H, J )
6.8, 4.0 Hz), 4.38 (dd, 1 H, J ) 7.2, 2.4 Hz), 4.02 (dd, 1 H, J )
4.0, 1.2 Hz), 3.35 (s, 3 H), 0.944 (s, 9 H), 0.930 (s, 9 H), 0.213 (2,
3 H), 0.152 (s, 3 H), 0.122 (s, 6 H). 13C NMR (100 MHz, CDCl3)
δ: 165.3, 133.0, 131.7, 129.9, 129.7, 128.4, 128.2, 128.1, 122.2,
108.2, 86.9, 83.3, 82.9, 73.0, 66.0, 55.2, 26.0, 25.9, 18.3, 18.2, -4.0,
-4.3, -4.4, -4.6. FTIR (thin film) cm-1: 2941 (s), 2862 (m), 1727
(s), 1477 (w), 1380 (w), 1268 (s), 1168 (s), 1095 (s), 877 (w), 842
(m). HRMS (MALDI-FTMS) m/z: calcd for C33 H48 O6 Si2 (M+
+ Na), 619.2881; found, 619.2876.
1
EtOAc/hexanes): 0.25. H NMR (400 MHz, CDCl3) δ: 7.27 (m,
5 H), 7.15 (m, 1 H), 7.02 (t, 2 H, J ) 7.6 Hz), 6.97 (m, 2 H), 5.97
(app t, 1 H, J ) 2.4 Hz), 4.87 (d, 1 H, J ) 11.2 Hz), 4.874 (s, 1
H), 4.25 (dd, 1 H, J ) 9.2, 2.4 Hz), 4.12 (m, 2 H), 3.36 (s, 3 H),
3.31 (dd, 1 H, J ) 11.2, 3.2 Hz), 2.23 (s, 3 H). 13C NMR (100
MHz, CDCl3) δ: 169.7, 138.1, 135.2, 128.92, 128.90, 128.6, 128.4,
128.3, 128.2, 128.1, 127.9, 109.2, 82.9, 76.1, 74.2, 73.3, 71.5, 55.6,
50.0, 21.1. FTIR (thin film) cm-1: 3389 (m), 2900 (w), 1724 (s),
1241 (s), 1116 (m), 1019 (m), 976 (m). HRMS (MALDI-FTMS)
m/z: calcd for C22 H24 O6 Se (M+ + Na), 487.0630; found,
487.0642.
Allylic Acetate 24. Lithium aluminum hydride (0.940 g, 24.80
mmol, 3.00 equiv) was cooled to 0 °C and then suspended in THF
(100 mL). 23 (4.93 g, 8.30 mmol, 1 equiv) in THF (10 mL + 5
mL rinse) was added via syringe to the flask containing LiAlH4,
and the resulting mixture was warmed to room temperature and
then heated to 50 °C for 3.5 h. The flask was cooled to room
temperature, then to 0 °C, and the excess LAH was quenched by
the addition of EtOAc. Saturated aqueous NaHCO3 was added via
addition funnel dropwise. The resulting fine gray solid was filtered
through Celite and washing with EtOAc, and the filtrate was washed
with brine, dried (sodium sulfate), and concentrated to yield an
oil. The crude alcohol was dried by coevaporation with benzene
(2×) and then dissolved in CH2Cl2 and cooled to 0 °C. 4-DMAP
(0.050 g) was added, followed by triethylamine (approximately 3.0
equiv, 4.06 mL) and acetic anhydride (approximately 3.0 equiv,
2.34 mL). The reaction was warmed to room temperature, stirred
for 20 h, and then poured onto ice and saturated aqueous NaHCO3.
It was stirred until the ice melted, then extracted with CH2Cl2 (3×)
and washed with 1 M HCl, saturated aqueous NaHCO3, and brine
and then dried (sodium sulfate) and concentrated. Chromatography
(silica, 5% EtOAc/hexanes, 40 mm × 6.5 in.) gave 24 (3.64 g,
82%, two steps), contaminated with a small amount of benzyl
acetate. TLC Rf (10% EtOAc/hexanes): 0.31. 1H NMR (400 MHz,
CDCl3) δ: 7.32 (m, 10 H), 6.73 (d, 1 H, J ) 16.0 Hz), 6.24 (dd,
1 H, J ) 16.0, 8.0 Hz), 5.61 (dd, 1 H, J ) 8.0 Hz), 4.73 (s, 1 H),
4.23 (dd, 1 H, J ) 7.6, 2.8 Hz), 4.11 (dd, 1 H, J ) 7.6, 4.0 Hz),
3.90 (d, 1 H, J ) 3.6 Hz), 3.41 (s, 3 H), 2.14 (s, 3 H), 0.93 (s, 9
H), 0.92 (s, 3 H), 0.12 (s, 3 H), 0.11 (s, 3 H), 0.10 (s, 3 H), 0.08
Diacetate 27. Compound 26 (0.034 g, 0.073 mmol, 1 equiv)
was dried by coevaporation with benzene (2×) and then dissolved
in CH2Cl2 (1 mL) and cooled to 0 °C. A few crystals of DMAP
were added followed by triethylamine (0.021 mL, 0.147 mmol, 2.00
equiv) and acetic anhydride (0.010 mL, 0.11 mmol, 1.50 equiv).
After 50 min, the reaction was poured onto ice and saturated
aqueous NaHCO3, stirred until the ice melted, and then extracted
with CH2Cl2 (3×). It was washed with 1 M HCl, saturated aqueous
NaHCO3, and brine and then dried (sodium sulfate) and concen-
trated. 32 (0.034 g, 92%) was isolated as a white solid (mp 151-
154 °C). TLC Rf (50% EtOAc/hexanes): 0.44. 1H NMR (400 MHz,
CDCl3) δ: 7.28 (m, 5 H), 7.15 (t, 1 H, J ) 7.2 Hz), 7.02 (t, 2 H,
J ) 8.4 Hz), 6.94 (m, 2 H), 5.95 (app t, 1 H, J ) 2.8 Hz), 5.08 (d,
1 H, J ) 4.4 Hz), 4.85 (s, 1 H), 4.82 (d, 1 H, J ) 11.6 Hz), 4.28
(dd, 1 H, J ) 10.2, 4.4 Hz), 4.21 (dd, 1 H, J ) 10.2, 2.0 Hz), 3.36
(s, 3 H), 3.29 (dd, 1 H, J ) 11.0, 2.4 Hz), 2.24 (s, 3 H), 2.10 (s,
3 H). 13C NMR (100 MHz, CDCl3) δ: 170.0, 169.7, 138.1, 135.2,
128.8, 128.77, 128.29, 128.27, 128.0, 107.4, 82.8, 76.8, 74.3, 72.2,
71.2, 55.7, 50.3, 21.2, 21.1. FTIR (thin film) cm-1: 2926 (w), 1753
(s), 1372 (w), 1222 (s), 1111 (m), 1026 (m). HRMS (MALDI-
FTMS) m/z: calcd for C24 H26 O7 Se (M+ + Na), 529.0736; found,
529.0734.
2240 J. Org. Chem., Vol. 71, No. 6, 2006

Synthesis of the Bicyclic Core of Octosyl Acid A
Compound 28. Selenide 27 (0.113 g, 0.224 mmol, 1 equiv) was
dried by coevaporation with benzene (2×) and then dissolved in
toluene (5 mL) and added via syringe to a vessel containing AIBN
(0.010 g, 0.056 mmol, 0.25 equiv). Tris(trimethylsilyl)silane (0.345
mL, 1.12 mmol, 5.00 equiv) was added via syringe, and the resulting
solution was thoroughly degassed and then heated to 110 °C for
15 h. The solution was cooled to room temperature, concentrated,
and then chromatographed directly (silica, 30% EtOAc/hexanes,
25 mm × 4 in. column) to yield bisacetate 28 (0.070 g, 90%) as
an oil. TLC Rf (30% EtOAc/hexanes): 0.24. 1H NMR (400 MHz,
CDCl3) δ: 7.33 (m, 5 H), 5.61 (d, 1 H, J ) 2.4 Hz), 5.17 (d, 1 H,
J ) 4.0 Hz), 4.90 (s, 1 H), 4.87 (d, 1 H, J ) 2.8 Hz), 4.29 (dd, 1
H, J ) 10.0, 4.4 Hz), 4.11 (dd, 1 H, J ) 10.4, 2.4 Hz), 3.40 (s, 3
H), 2.20 (s, 3 H), 2.19 (s, 3 H), 2.14 (m, 1 H), 1.92 (m, 1 H). FTIR
(thin film) cm-1: 2942 (w), 1748 (s), 1378 (w), 1229 (s), 1107
(m), 1063 (m), 1030 (w). HRMS (MALDI-FTMS) m/z: calcd for
C18 H22 O7 (M+ + Na), 351.1438; found, 351.1448.
Acknowledgment. We thank Dr. Robert Chang for as-
sistance with NMR interpretation and the National Institutes of
Health for support of this research (NIGMS-GM60875; NIGMS-
GM07240 training grant support for J.D.M.).
Supporting Information Available: Detailed experimental
1
procedures for compounds 9-15 and copies of H NMR spectra
for compounds 9-15, 18, 20-27, 29, and 31. This material is
available free of charge on the Internet at http://pubs.acs.org.
JO052025S
J. Org. Chem, Vol. 71, No. 6, 2006 2241