Total synthesis of (-)-hennoxazole A

Article abstract of DOI:10.1021/jo7018015

(Chemical Equation Presented) An enantioselective, convergent total synthesis of the antiviral marine natural product (-)-hennoxazole A is completed in 14 steps (longest linear sequence) from commercially available 4-methyloxazole-2-carboxylic acid. Synth

Full text of DOI:10.1021/jo7018015

Total Synthesis of (-)-Hennoxazole A
Thomas E. Smith,* Wen-Hsin Kuo, Emily P. Balskus, Victoria D. Bock, Jennifer L. Roizen,
Ashleigh B. Theberge, Kathleen A. Carroll, Tomoki Kurihara, and Jeffrey D. Wessler
Department of Chemistry, Williams College, Williamstown, Massachusetts 01267
tsmith@williams.edu
ReceiVed August 22, 2007
An enantioselective, convergent total synthesis of the antiviral marine natural product (-)-hennoxazole
A is completed in 14 steps (longest linear sequence) from commercially available 4-methyloxazole-2-
carboxylic acid. Synthesis of the C1-C15 pyran/bisoxazole fragment takes advantage of an aldol-like
coupling between a dimethyl acetal and an N-acetylthiazolidinethione for the direct, stereoselective
installation of the C8-methoxy-bearing stereocenter. A one-pot acetoacetate acylation/decarboxylation/
cyclodehydration of another elaborate thiazolidinethione allows for rapid assembly of the pyran-based
ring system. Synthesis of the C15-C25 skipped triene side chain fragment makes use of a [2,3]-Wittig-
Still rearrangement for efficient installation of the trisubstituted Z-double bond. Key late-stage coupling
of the two fragments is effected by deprotonation of the methyl group on the bisoxazole system using
lithium diethylamide, followed by alkylation with an allylic bromide side chain segment to form the
C15-C16 bond.
Introduction
mides4 are the only other natural products known to include
two contiguous 2,4-disubstituted oxazoles in their frameworks.5
Other structurally distinctive features of the hennoxazoles
include the functionalized pyranyl ring and the nonconjugated
triene side chain containing a trisubstituted Z-double bond and
a remote stereogenic center. The original structure determination
work established the correct atomic connectivity but left many
of the stereochemical details ambiguous; although the assign-
ment of relative stereochemistry about the C2-C6 pyran ring
In 1991, a research team led by Paul Scheuer at the University
of Hawaii reported the isolation and structural elucidation of
hennoxazoles A-D (Figure 1) from a species of Polyfibro-
spongia sponge off the coast of Miyako island in Okinawa,
Japan.1 Further investigation by Higa2 disclosed four additional
members of this natural product family, hennoxazoles E-G and
hennoxazole A acetate. The most abundant member of the
group, hennoxazole A, is also the most active, displaying
antiviral activity against herpes simplex 1 (IC50 ) 0.6 µg/mL)
and peripheral analgesic activity comparable to that of in-
domethacin.
(4) For a lead reference on the diazonamides see: Poriel, C.; Lachia,
M.; Wilson, C.; Davies, J. R.; Moody, C. J. J. Org. Chem. 2007, 72, 2978-
2987.
Structurally, the hennoxazoles contain a bisoxazole ring
system at their molecular core. The muscorides3 and diazona-
(5) Several other natural products containing more than two contiguous
oxazoles are also known, including the mycalolides, ulapualides, and
kabiramidesswhich have three directly linked oxazole ringssand telom-
estatin, which has seven. For a review of the syntheses of oxazole-containing
natural products, see: (a) Yeh, V. S. C. Tetrahedron 2004, 60, 11995-
12042. (b) Palmer, D., Ed. Heterocyclic Compounds; J. Wiley and Sons:
New York, 2003; Vol. 60, Part A. (c) Riego, E.; Herna´ndez, D.; Albericio,
F.; AÄ lvarez, M. Synthesis 2005, 1907-1922. For a lead reference on
telomestatin, see: (d) Doi, T.; Yoshida, M.; Shin-ya, K.; Takahashi, T. Org.
Lett. 2006, 8, 4165-4167.
(1) Ichiba, T.; Yoshida, W. Y.; Scheuer, P. J.; Higa, T.; Gravalos, D. G.
J. Am. Chem. Soc. 1991, 113, 3173-3174.
(2) Higa, T.; Tanaka, J.-i.; Kitamura, A.; Koyama, T.; Takahashi, M.;
Uchida, T. Pure Appl. Chem. 1994, 66, 2227-2230.
(3) For a lead reference on the muscorides see: Coqueron, P.-Y.; Didier,
C.; Ciufolini, M. A. Angew. Chem., Int. Ed. 2003, 42, 1411-1414.
10.1021/jo7018015 CCC: $40.75 © 2008 American Chemical Society
Published on Web 12/06/2007
142
J. Org. Chem. 2008, 73, 142-150

Total Synthesis of (-)-Hennoxazole A
The same nonbonding interactions that lead to a helical second-
ary structure in this portion of the molecule may serve to inhibit
undesirable isomerization of the side chain double bonds.
Since Wipf’s ground-breaking synthesis, two other total
syntheses of hennoxazole A, as well as a number of synthetic
studies, have been reported.7 The second total synthesis, by
Williams and co-workers, took advantage of a directed allylbo-
rane addition to an aldehyde to form the C7-C8 bond in the
key fragment coupling step.8 The trisubstituted Z-olefin stere-
ochemistry was secured via syn-elimination of a 1,2-diol
derivative. Also notable is the fact that the bisoxazole unit was
fabricated early in the route and carried through the entire
sequence. The Williams synthesis proceeded in a 20-step LLS
from sodium γ-hydroxybutyrate. The third total synthesis, by
Shioiri and co-workers, shared Wipf’s strategy for late-stage
oxazole formation as the key coupling step.9 The stereochemistry
of the trisubstituted olefin was secured by a Z-selective Still-
Horner-Wadsworth-Emmons reaction. This synthesis was
accomplished in a 25-step LLS from (R)-glycidyl tosylate.
Herein, we detail our own investigations which have culminated
in a 14-step LLS total synthesis of (-)-hennoxazole A.10
Results and Discussion
Retrosynthetic Analysis. The development of relatively mild
conditions11 for the preparation of oxazoles has made the late-
stage assembly of these ring systems a common, albeit not
always efficient, strategy in the synthesis of oxazole-containing
targets. Approaches involving end game functionalization of
intact oxazole rings, however, provide the opportunity to use
relatively simple oxazoles as starting materials and then carry
these practically inert heterocycles through a variety of synthetic
transformations unscathed.12 In consideration of these issues,
our synthetic plan for hennoxazole A (Scheme 1) diverges from
precedent and involves late-stage construction of the C15-C16
bond by deprotonation of an elaborate bisoxazole fragment (4)
at the C15-methyl group, followed by alkylation with an allylic
bromide side chain fragment (5). Further disconnection leads
back to achiral bisoxazole dimethyl acetal 6 as a precursor to
the direct stereoselective installation of the C8-methoxy sub-
stituent. Likewise, allylic alcohol 7 could be used as the
progenitor of the isolated C22 stereocenter.
FIGURE 1. Bisoxazole natural products.
was secure, the relative stereochemistry of the distant C8 and
C22 stereogenic centers and the absolute configuration remained
unknown. Pioneering synthetic work by Wipf led to the total
synthesis of the unnatural antipode of hennoxazole A and solved
the stereochemical puzzle.6 Key to the Wipf synthesis was a
late-stage oxazole ring assembly that involved fragment coupling
by amide formation followed by cyclodehydration. The Z-
geometry of the trisubstituted alkene in the side chain was
established through a vinylstannane reagent that was stereose-
lectively prepared early on. This landmark synthesis proceeded
in a 28-step longest linear sequence (LLS) from mercaptoimi-
dazole. Also discovered in the course of this synthesis was the
important fact that hennoxazole A and its C22 epimer are
chromatographically and spectroscopically indistinguishables
except by optical rotation. The consequences of this detail were
significant in that any other planned syntheses would require
the installation of C22 with perfect stereochemical fidelity, since
separation would not be possible at the end of the synthesis.
Wipf’s work also suggested that the skipped triene side chain
adopts an interesting helical conformation, principally by virtue
of A1,3 strain imparted by the trisubstitued Z-double bond, and
that this structural feature may contribute to the biological
activity of hennoxazole A. This conformational predisposition
may also have synthetic ramifications that help to mollify con-
cerns about the stability of the nonconjugated triene side chain.
(7) For other synthetic studies see: (a) Barrett, A. G. M.; Kohrt, J. T.
Synlett 1995, 415-416. (b) Vakalopoulos, A.; Hoffmann, H. M. R. Org.
Lett. 2001, 3, 177-180. (c) Chen, Y. K.; Walsh, P. J. Am. Chem. Soc.
2004, 126, 3702-3703. (d) Sakakura, A.; Kondo, R.; Ishihara, K. Org.
Lett. 2005, 7, 1971-1974. (e) Zylstra, E. J.; She, M. W.-L.; Salamant, W.
A.; Leahy, J. W. Synlett 2007, 623-627.
(8) Williams, D. R.; Brooks, D. A.; Berliner, M. A. J. Am. Chem. Soc.
1999, 121, 4924-4925.
(9) (a) Cheng, Z. Z.; Hamada, Y.; Shioiri, T. Synlett 1997, 109-110.
(b) Shioiri, T.; McFarlane, N.; Hamada, Y. Heterocycles 1998, 47, 73-76.
(c) Yokokawa, F.; Asano, T.; Shioiri, T. Org. Lett. 2000, 2, 4169-4172.
(d) Yokokawa, F.; Asano, T.; Shioiri, T. Tetrahedron 2001, 57, 6311-
6327.
(10) Portions of this work have been reported in previous papers: (a)
Smith, T. E.; Balskus, E. P. Heterocycles 2002, 57, 1219-1225. (b) Smith,
T. E.; Kuo, W.-S.; Bock, V. D.; Roizen, J. L.; Balskus, E. P.; Theberge, A.
B. Org. Lett. 2007, 9, 1153-1155.
(11) (a) Phillips, A. J.; Uto, Y.; Wipf, P.; Reno, M. J.; Williams, D. R.
Org. Lett. 2000, 2, 1165-1168. (b) Williams, D. R.; Lowder, P. D.; Gu,
Y.-G.; Brooks, D. A. Tetrahedron Lett. 1997, 38, 331-334. (c) Sakakura,
A.; Kondo, R.; Ishihara, K. Org. Lett. 2005, 7, 1971-1974. (d) Wipf, P.;
Aoyama, Y.; Benedum, T. E. Org. Lett. 2004, 6, 3593-3595.
(12) For a review of the uses of oxazoles as protected carboxylate
equivalents see: (a) Wasserman, H. H.; McCarthy, K. E.; Prowse, K. S.
Chem. ReV. 1986, 86, 845-856. See also: (b) Evans, D. A.; Nagorny, P.;
Xu, R. Org. Lett. 2006, 8, 5669-5671.
(6) (a) Wipf, P.; Lim, S. J. Am. Chem. Soc. 1995, 117, 558-559. (b)
Wipf, P.; Lim, S. Chimia 1996, 50, 157-167.
J. Org. Chem, Vol. 73, No. 1, 2008 143

Smith et al.
SCHEME 1. Synthetic Strategy toward Hennoxazole A
SCHEME 3. Preparation of the Bisoxazole Core
SCHEME 2. Oxazole and Bisoxazole Alkylation
Background
azole (10) using n-BuLi at -78 °C followed by alkylation with
methyl iodide gives a 14:86 ratio of products (13:14) favoring
ring methylation, while the use of LiNEt2 leads to alkylation
solely at the C2-methyl site. This reversal of regioselectivity is
thought to arise from the ability of diethylamine to mediate the
low-temperature equilibration of a kinetic mixture of otherwise
noninterconverting lithiated intermediates (11 and 12).17 Given
these results, we thought it possible that the undesired C13-
alkylation observed by Williams might be a consequence of
kinetic (ring) deprotonation and that we might achieve the
desired C15-alkylation on a bisoxazole system such as 6 by equi-
librating to the thermodynamic (methyl) lithiated intermediate.
To test the viability of our key side chain coupling strategy,
we set out to prepare bisoxazole 23 as a model substrate
(Scheme 3). Commercially available (()-serine methyl ester
hydrochloride (15) was condensed with ethyl acetimidate (16)
to give oxazoline 17, which was oxidized to oxazole 18 as
described in Leahy’s synthesis of rhizoxin D.18 Next, following
Pattenden’s preparation,19 methyl ester 18 was saponified to
carboxylic acid 19, which was activated as its acid chloride and
combined with another equivalent of serine methyl ester to give
amide 20. It is interesting to note that both ester 18 and acid 19
are now commercially available. The one-pot cyclodehydration/
Oxazole Alkylation Studies. Synthetically useful lateral
metalations of some 2-methyloxazole and thiazole systems have
been reported.13 If these rings are unsubstituted at C5 (C13,
hennoxazole numbering), competitive deprotonation of the C5-
ring hydrogen is frequently observed (Scheme 2).14 In fact,
Williams established that bisoxazole 8 is lithiated exclusively
at the C13-position with n-BuLi,15 suggesting that alkylation of
4 at C15 might be problematic when R ) H. Despite this result,
previous work in confronting a similar problem during the
synthesis of phorboxazole demonstrated that the regioselectivity
of some oxazole deprotonations can be inverted by the choice
of base.16 For example, deprotonation of 2-methyl-4-phenylox-
(13) Lipshutz, B. H.; Hungate, R. W. J. Org. Chem. 1981, 46, 1410-
1413. (b) Whitney, S. E.; Rickborn, B. J. Org. Chem. 1991, 56, 3058-
3063. (c) Vedejs, E.; Zajac, M. A. Org. Lett. 2001, 3, 2451-2454. (d)
Entwistle, D. A.; Jordan, S. I.; Montgomery, J.; Pattenden, G. Synthesis
1998, 603-612. (e) Garey, D; Ramirez, M.-l.; Gonzales, S.; Wertsching,
A.; Tith, S.; Keefe, K.; Pena, M. R. J. Org. Chem. 1996, 61, 4853-4856.
(14) (a) Knaus, G.; Meyers, A. I. J. Org. Chem. 1974, 39, 1192-1195.
(b) Meyers, A. I; Lawson, J. P. Tetrahedron Lett. 1981, 22, 3163-3166.
(c) Meyers, A. I.; Walker, D. G. J. Org. Chem. 1982, 47, 2999-3000. (d)
Meyers, A. I.; Lawson, J. P.; Walker, D. G.; Linderman, R. J. J. Org. Chem.
1986, 51, 5111-5123. (e) Hamana, H.; Sugasawa, T. Chem. Lett. 1983,
333-336. For a review of oxazole metalation see: (f) Iddon, B. Heterocycles
1994, 37, 1321-1346.
(16) For the original solution to this problem see: (a) Evans, D. A.;
Cee, V. J.; Smith, T. E.; Santiago, K. J. Org. Lett. 1999, 1, 87-90. (b)
Evans, D. A.; Cee, V. J.; Smith, T. E.; Fitch, D. M.; Cho, P. S. Angew.
Chem., Int. Ed. 2000, 39, 2533-2536. (c) Evans, D. A.; Fitch, D. M.; Smith,
T. E.; Cee, V. J. J. Am. Chem. Soc. 2000, 122, 10033-10046. For
subsequent applications of this method to the synthesis of phorboxazoles
see: (d) Gonza´lez, M. A.; Pattenden G. Angew. Chem., Int. Ed. 2003, 42,
1255-1258. (e) Li, D.-R.; Zhang, D.-H.; Sun, C.-Y.; Zhang, J.-W.; Yang,
L.; Chen, J.; Liu, B.; Su, C.; Zhou, W.-S.; Lin, G.-Q. Chem. Eur. J. 2006,
12, 1185-1204. (f) White, J. D.; Kuntiyong, P.; Lee, T. H. Org. Lett. 2006,
8, 6039-6042
(17) For further investigation of this effect see: Smith, T. E.; Mourad,
M. S.; Velander, A. J. Heterocycles 2002, 57, 1211-1217.
(15) Williams, D. R.; Brooks, D.; Meyer, K. G. Tetrahedron Lett. 1998,
39, 8023-8026.
(18) Lafontaine, J. A.; Provencal, D. P.; Gardelli, C.; Leahy, J. W. J.
Org. Chem. 2003, 68, 4215-4234.
144 J. Org. Chem., Vol. 73, No. 1, 2008

Total Synthesis of (-)-Hennoxazole A
SCHEME 4. Bisoxazole Model Alkylations
known allylic alcohol 7 (Scheme 6). For our initial studies, we
prepared this material according to Williams’ published route,8
which derives from (S)-3-pentyn-2-ol (31). This compound can
be obtained from the C-methylation of commercially available
(S)-3-butyn-2-ol or by enzymatic resolution of racemic 31.24
Interestingly, we discovered some erosion of enantiomeric
excess in our synthetic sequence resulting from the [2,3]-Wittig
rearrangement.25 Although there was essentially complete
transfer of stereochemical information between methallyl ether
32 and major syn-product 7, it turned out that the same was not
true for minor anti-product 33, which was obtained in only 77%
ee.26 Since both diastereomers converge on the same compound
(37; see Scheme 7), any ent-33 carried forward would be
transformed into ent-37, thus eroding the side chain ee. The
importance of side chain enantiomeric purity to the eventual
success of the asymmetric synthesis has already been noted.
We were able to remove the offending anti-diastereomer (33)
via chromatography on silver nitrate-impregnated silica gel,27
providing syn-diastereomer 7 in 95% ee.
1
a Only product observed by H NMR. No yield calculated.
oxidation method of Wipf and Williams11a was then used to
generate bisoxazole methyl ester 21 in 82% yield, an improve-
ment over previous multistep preparations.19,11b Partial reduction
with DIBAL-H gave aldehyde 22 in quantitative yield. Finally,
Noyori conditions20 were used to generate dimethyl acetal 23
in 96% yield.
We anticipated that kinetic resolution of racemic 7 using Fu’s
planar-chiral DMAP catalyst could be used as an alternative
asymmetric preparation of this starting material.28 Similar allylic
alcohols have been resolved with selectivity factors in the 10-
25 range.29 In the event, kinetic resolution of rac-7 using acetic
anhydride proceeded with good selectivity (s ) 10.4), but the
reaction was too slow to be of practical use. Experiments at
higher temperatures or catalyst/reagent loadings were not
investigated.
Finally, we investigated the asymmetric pentenylation method
of Hoffmann.30 Allyl boronate 35 was prepared and allowed to
react with methacrolein (36). This very direct approach provided
desired allylic alcohol 7 in 81% yield and 98% ee.
With a supply of enantiopure starting material 7 in hand, we
proceeded with our synthesis of side chain 5 (Scheme 7). The
trisubstituted C20-C21 Z-double bond was expeditiously installed
via a [2,3]-Wittig-Still rearrangement; avoidance of A1,2-strain
in the early transition state presents one diastereoface of the
C19-C20 double bond to the intermediate C18-anion, leading to
homoallylic alcohol 37 as a single isomer.31 Notably, this key
transformation accomplishes in one step what had required six
steps in Williams’ synthesis.32 Dess-Martin oxidation33 next
gave aldehyde 38, which was subjected to an E-selective
Horner-Wadsworth-Emmons olefination to give ester 39.
Reduction with DIBAL-H then provided allylic alcohol 40 in
Consistent with Williams’ studies of 8,15 treatment of model
bisoxazole 23 with n-BuLi led to deprotonation exclusively at
the C13-ring position (Scheme 4), with no deprotonation occurr-
ing at the C15-methyl group. Unfortunately, replacing the base
with LDA or LiNEt2 did not alter the regioselectivity, suggesting
that, in the case of 23, deprotonation at C13 is thermodynamically
as well as kinetically favored.21 Attempts to alkylate the dianion
of 23 were not fruitful. As an indirect solution to this
predicament, we chose to block C13 with a silyl protecting
group.22 Although TMS and TES groups were found to be too
labile for this purpose, since they underwent silyl transfer
processes upon C15-deprotonation, the TBS group proved to be
suitable. Treatment of 23 with n-BuLi followed by addition of
TBSOTf led to C13-protected bisoxazole 24b in 88% yield.
To model our key coupling step, we treated protected
bisoxazole 24b with several different strong bases and quenched
with MeI (Scheme 5). Gratifyingly, alkylation occurred pre-
dominantly at the desired C15 site to give 25, with LiNEt2
providing the best results. It is interesting to note that n-BuLi
and LDA both gave poor conversion and small amounts of
product 26smethylated at both the C15-methyl and C10-ring
positionssat the expense of starting material conversion.23 No
significant monomethylation at C10 was observed. To better
mimic the reactivity of the actual side chain fragment (5), we
also alkylated 24b with allyl iodide and prenyl bromide, both
of which gave excellent results with LiNEt2. Importantly,
treatment of the alkylated products (27 and 29) with TBAF
demonstrated that the oxazole could be cleanly deprotected
under mild conditions.
(24) (a) Lowe, J. T.; Panek, J. S. Org. Lett. 2005, 7, 1529-1532. (b)
Lowe, J. T.; Panek, J. S. Org. Lett. 2007, 9, 2753.
(25) Tsai, D. J.-S.; Midland, M. M. J. Am. Chem. Soc. 1985, 107, 3915-
3918.
(26) Stereochemical erosion in the syn-diastereomer of this type of [2,3]-
Wittig rearrangment has been observed previously: Tsubuki, M.; Kamata,
T.; Nakatani, M.; Yamazaki, K.; Matsui, T.; Honda, T. Tetrahedron:
Asymmetry 2000, 11, 4725-4736.
(27) Williams, C. M.; Mander, L. N. Tetrahedron 2001, 57, 425-447.
(28) For a review see: Fu, G. C. Acc. Chem. Res. 2004, 37, 542-547.
(29) Bellemin-Laponnaz, S.; Tweddell, J.; Ruble, J. C.; Breitling, F. M.;
Fu, G. C. Chem. Commun. 2000, 1009-1010.
(30) For the original reference, see: (a) Hoffmann, R. W.; Ditrich, K.;
Ko¨ster, G.; Stu¨rmer, R. Chem. Ber. 1989, 122, 1783-1789. For an improved
preparation of the (Z)-pentenyl boronate reagent, see: (b) Bahnck, K. B.;
Rychnovsky, S. D. Chem. Commun. 2006, 2388-2390.
(31) (a) Still, W. C.; Mitra, A. J. Am. Chem. Soc. 1978, 100, 1927-
1928. (b) Still, W. C.; McDonald, J. H., III; Collum, D. B.; Mitra, A.
Tetrahedron Lett. 1979, 20, 593-594. For a review see: (c) Mikami, K.;
Nakai, T. Synthesis 1991, 594-604.
(32) Homoallylic alcohol 37 is also an intermediate in Williams’s
synthesis, but required six steps to prepare from 7.
Side Chain Fragment Synthesis. Having established a viable
coupling strategy, we set out to prepare the actual skipped triene
side chain fragment (5). As our starting material, we selected
(19) Chattopadhyay, S. K.; Kempson, J.; McNeil, A.; Pattenden, G.;
Reader, M.; Rippon, D. E.; Waite, D. J. Chem. Soc., Perkin Trans. 1 2000,
2415-2428.
(20) Tsunoda, T.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980, 21,
1357-1358.
(21) Williams carried out semiempirical calculations also suggesting that
the ring-lithiated intermediate was thermodynamically favored due to
chelation with the adjacent oxazole nitrogen. See ref 15.
(22) Silylation at C5 was one approach to solving “the oxazole problem”
in virginiamycin: (a) Wood, R. D.; Ganem, B. Tetrahedron Lett. 1983, 24,
4391-4392. (b) Fujita, E. Heterocycles 1984, 21, 41-60.
(23) Product identities and ratios were determined using a combination
1
of H NMR and GC-MS analysis.
(33) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155-4156.
J. Org. Chem, Vol. 73, No. 1, 2008 145

Smith et al.
SCHEME 5. Protected Bisoxazole Model Alkylations
SCHEME 6. Preparation of the Side Chain Precursor (7)
SCHEME 7. Synthesis of the C16-C25 Side Chain Fragment
(5)
SCHEME 8. Model Fragment Coupling with the Actual
Side Chain (5)
76% yield over the three steps. Treatment with mesyl chloride
and lithium bromide completed the synthesis of electrophilic
coupling fragment 5.
Pyran/Bisoxazole Fragment Synthesis. Encouraged by these
experiments, we carried on with our synthesis of pyran/
bisoxazole fragment 4. The first significant challenge in this
endeavor involved establishing the desired absolute stereochem-
ical configuration at C8. It was our hope that dimethyl acetal
24b could be used to install the C8-methoxy group directly,
without requiring the formation of a methyl ether via a multiple-
step reaction sequence. The stereoselective aldol-like reactions
Once the actual side chain (5) was prepared, a final model
coupling was undertaken (Scheme 8). Deprotonation of TBS-
protected model bisoxazole 24b with LiNEt2, as before, and
treatment with side chain allylic bromide 5 gave the desired
alkylation product (41) with no evidence of a competing SN2′
pathway. This comforting result indicated an even greater likeli-
hood that our final fragment coupling event would be successful.
146 J. Org. Chem., Vol. 73, No. 1, 2008

Total Synthesis of (-)-Hennoxazole A
SCHEME 9. Condensation Reactions of N-Acetylthiazolidinethiones with Dimethyl Acetals
a Combined yield of both diastereomers. b Reaction only proceeded to 50% conversion. c No yield calculated. d Acetal was consumed, but 47b and 48b
not observed.
between N-acylthiazolidinethiones and acetals reported by Urp´ı
bisoxazole dimethyl acetal 24b (entry 3), the opposite diaste-
reomer was obtained preferentially. We posit that this reversal
in selectivity may be due to the ability of an oxazole nitrogen
atom to coordinate with the titanium center and alter the
organization of the transition state. Regardless of the rationale,
the two diastereomers were produced in a 20:80 ratioscertainly
acceptable if we were to use the opposite enantiomer of the
thione auxiliary. Unfortunately, the two products (47a and 48a)
were very difficult to separate, which limited the practical
application of this method to our synthesis. As a consequence,
we probed additional reaction conditions.
provided inspiration toward this end.34
Several chiral auxiliaries and soft enolization conditions were
surveyed to find suitable reaction conditions (Scheme 9). Urp´ı’s
original conditions were explored first (entry 1). Treatment of
the titanium enolate of i-Pr-substituted thiazolidinethione 44a
with benzaldehyde dimethyl acetal (43) in the presence of the
Lewis acid, BF3‚OEt2, gave anti-isomer 45a and syn-isomer 46a
in an 85:15 ratio in accordance with Urp´ı’s reported results.
The diastereoselectivity for this reaction can be rationalized by
an open transition state, such as 49, that allows for an anti-
approach of the enolate to the activated carbonyl and minimizes
interaction between the bulky R and thione substituents.
Interestingly, when these same conditions were used with
Intrigued by Sammakia’s recent reports of boron-mediated
aldol reactions of N-acetylyhiazolidinethiones,35 we decided to
evaluate whether these conditions might be suitable for reactions
with acetals. In the simplest case (entry 2) this seemed to be
borne out, although the diastereomer ratio and isolated yield
were suboptimal. Importantly, the bisoxazole substrate (entry
(34) Cosp, A.; Romea, P.; Urp´ı, F.; Vilarrasa, J. Tetrahedron Lett. 2001,
42, 4629-4631.
J. Org. Chem, Vol. 73, No. 1, 2008 147

Smith et al.
SCHEME 10. Synthesis of the C1-C15 Pyran/Bisoxazole Fragment (4a) and Completion of the Total Synthesis
4) gave the same sense of asymmetric induction as the simple
benzaldehyde dimethyl acetal. A transition-state model for this
boron-mediated transformation (50) is similar to that of the
titanium-promoted case (49), but differs critically in the facility
of the titanium center to coordinate with additional ligands. Since
Sammakia demonstrated much higher levels of diastereoselec-
tion when he substituted a t-Bu substituent (or the pseudo-t-Bu
group in the form of a -CMe2(OTES) group) for the i-Pr
substituent on the chiral auxiliary, we explored those changes
as well using 44b (entries 6 and 8). Greater diastereoselectivity
was indeed observed. Under these conditions, the desired major
isomer 47b was obtained in 53% isolated yield, and although
this yield was somewhat lower than could be obtained for 47a,
the products were easier to separate.36 About 10% â-eliminated
product was also obtained under these conditions. The higher
susceptibility of the -CMe2(OTES)-substituted auxiliaries
toward hydrolysis on silica gel may also be reflected in the
somewhat diminished isolated yields obtained after chromatog-
raphy. Additionally, a more conventional two-step approach
involving an auxiliary-directed aldol reaction followed by methyl
ether formation was thwarted by â-elimination difficulties.
Overall, this direct pathway (entry 8) appeared to be the best
compromise of increased structural complexity and yield and
was appropriated to advance material to the forefront of the
synthesis.
was obtained with the bisoxazole acetal (entry 7). It was also
demonstrated that the mesityl-substituted auxiliaries (such as
44c), very recently disclosed by Crimmins,37 do not provide
useful levels of diastereoselectivity for these types of condensa-
tion reactions (entries 9 and 10).
Pressing on with our synthesis (Scheme 10), the thione
auxiliary was reductively cleaved from 47b using DIBAL-H.38
The resulting â-methoxy aldehyde was prone to â-elimination
and thus was immediately subjected to an auxiliary-directed
acetate aldol reaction using ent-44a. Reagent-based stereochem-
ical control was required to set the desired 1,3-syn-relationship
between the C6 and C8 oxygen substituents, effectively over-
riding the 1,3-anti-selectivity bias intrinsic to such â-methoxy
aldehyde substrates.39 We eventually settled on the original tin
enolate conditions of Nagao for this transformation to give aldol
adduct 51.40 Use of Sammakia’s auxiliary (44b) and boron
enolate gave a lower selectivity (67:34) and isolated yield (43%
yield of an inseparable mixture of product diastereomers). We
expect that Crimmins’ mesityl auxiliary (ent-44c) and titanium
enolate would be best suited here, but have not evaluated this
prospect.
Encouraged by our previous success with the kavalactones
in efficiently refunctionalizing N-acylthiazolidinethiones,41 we
postulated that this similar aldol product (51) might be directly
converted into methyl ketone 52. Gratifyingly, treatment with
magnesium acetoacetate42 and imidazole not only led to the
desired transacylation reactionspresumably through the inter-
mediacy of an acyl imidazolidesbut also, upon quenching with
In further model studies, the titanium enolization conditions
did not appear to be entirely compatible with the -CMe2-
(OTES)-substituted systems. Poor conversion was observed with
benzaldehyde dimethyl acetal (entry 5), and no desired product
(37) Crimmins, M. T.; Shamszad, M. Org. Lett. 2007, 9, 149-152.
(38) (a) Crimmins, M. T.; King, B. W.; Tabet, E. A.; Chaudhary, K. J.
Org. Chem. 2001, 66, 894-902. (b) Crimmins, M. T.; Slade, D. J. Org.
Lett. 2006, 8, 2191-2194.
(39) Evans, D. A.; Dart, M. J.; Duffy, J. L.; Yang, M. G. J. Am. Chem.
Soc. 1996, 118, 4322-4343.
(40) Nagao, Y.; Hagiwara, Y.; Kumagai, T.; Ochiai, M.; Inoue, T.;
Hashimoto, K.; Fujita, E. J. Org. Chem. 1986, 51, 2391-2393.
(41) Smith, T. E.; Djang, M.; Velander, A. J.; Downey, C. W.; Carroll,
K. A.; van Alphen, S. Org. Lett. 2004, 6, 2317-2320.
(35) (a) Zhang, Y.; Phillips, A. J.; Sammakia, T. Org. Lett. 2004, 6,
23-25. (b) Zhang, Y.; Sammakia, T. Org. Lett. 2004, 6, 3139-3141. (c)
Zhang, Y.; Sammakia, T. J. Org. Chem. 2006, 71, 6262-6265.
(36) As an additional benefit, product 47b could be reduced to the
corresponding aldehyde in a single step using DIBAL-H, whereas attempted
reduction of 47a and 48a always gave significant amounts of overreduction
to the primary alcohols, which would necessitate a two-step reduction/
oxidation sequence.
148 J. Org. Chem., Vol. 73, No. 1, 2008

Total Synthesis of (-)-Hennoxazole A
cm-1; [R]D24 ) +5.4° (c ) 1.01, CHCl3); HRMS (EI) exact mass
calcd for C9H16O [M]+ 140.1201, found 140.1192.
aqueous acid, provided cyclodehydrated pyrone 53. Notably,
this transformation is accomplished in a single step without the
need for protecting groups and made early approaches involving
elaboration of a C2-C6 δ-lactone obsolete. Luche reduction43
gave an unstable allylic alcohol, which was immediately
protected as its TBS ether (54) in a combined yield of 61%.
Mixed methyl acetal formation, in accordance with Wipf’s
pioneering work,6 finalized the synthesis of C1-C15 pyran/
bisoxazole coupling fragment 4a.
With both coupling partners in hand, we endeavored to unite
them in the pivotal transformation of the synthesis. Using the
same conditions that had been successful for model fragment
couplings, treatment of bisoxazole 4a with LiNEt2 and allylic
bromide 5 led to the anticipated union in 77% yield along with
7% unalkylated 4a, thus securing the entire carbon framework
of the natural product. Double TBS deprotection of 55 provided
(-)-hennoxazole A, having spectra identical to those published
previously.6,8,9
Preparation of Trisubstituted Alkene 37.32 To a 10 mL
concentration flask was added KH (290 mg of a 35% dispersion in
mineral oil, 2.38 mmol, 2.2 equiv). The white solid was washed in
dry pentanes (3 × 5 mL) and suspended in dry THF (2 mL) at
0 °C. To this suspension was added allylic alcohol 7 (152 mg, 1.08
mmol, 1 equiv) via cannula in THF (1.2 mL), producing a cloudy
yellow solution. Dibenzo-18-crown-6 (catalytic amount) was added
as a white powder, followed by addition of Bu3SnCH2I44 (560 mg,
1.30 mmol, 1.2 equiv) via cannula in THF (0.8 mL). The solution
became a light brown color and was warmed to rt. After 2 h, the
solution was cooled to -78 °C. n-Butyllithium (1.1 mL of a 1.56
M in hexanes, 1.63 mmol, 1.5 equiv) was added dropwise via
syringe down the cold wall of the flask, forming an orange solution.
The solution was stirred at -78 °C for 2 h. The orange color
gradually changed to light brown and then became more yellow.
The reaction mixture was partitioned between Et2O (10 mL) and
satd aq NH4Cl (10 mL), and the layers were separated. The aqueous
layer was extracted with Et2O (2 × 10 mL), and the combined
organic layers were dried over Na2SO4, filtered, and concentrated
in vacuo to yield a clear, yellow oil and a red-brown precipitate.
The product was purified via automated silica column chromatog-
raphy (0 f 15% EtOAc/hexanes over 30 min, 110 g column; TLC
Rf ) 0.42, 15% EtOAc/hexanes, anisaldehyde stain) to provide
homoallylic alcohol 37 (138 mg, 83% yield) as a clear, colorless
oil. Due to the volatility of alcohol 37, it should not be exposed to
pressures lower than 5 mmHg. Chiral GC (Chiraldex G-TA column,
30 m × 0.25 mm, 30 psi of He, 50 °C, 0.6 °C/min ramp) gave
only 85% resolution of the two enantiomer peaks but was consistent
Conclusions
We have developed a convergent asymmetric route to
hennoxazole A that proceeds in a longest linear sequence of 17
steps from serine methyl ester and only 14 steps from com-
mercially available 4-methyloxazole-2-carboxylic acid (19).
Contributing to the efficiency of this synthesis is the direct use
of a dimethyl acetal for the stereocontrolled installation of the
C8-methyl ether in a single step. The rapid functionalization of
the thiazolidinethione-derived aldol adducts and the necessity
of only two (TBS) protecting groupsswhich are both removed
in the final stepsalso convey significant step economy. Finally,
the key fragment coupling, featuring the alkylation of an intact
bisoxazole core, renders this synthesis highly convergent.
with a product ee identical to that of the starting material (tmajor
)
45.37 min, tminor ) 45.99 min). Other data: 1H NMR (500 MHz,
CDCl3) δ 5.45-5.34 (m, 2H), 5.18 (d, J ) 9.6 Hz, 1H), 3.72-
3.62 (m, 2H), 3.12-3.03 (m, 1H), 2.37 (dt, J ) 13.4, 6.4 Hz, 1H),
2.30 (dt, J ) 13.4, 6.4 Hz, 1H), 1.72 (d, J ) 1.2 Hz, 3H), 1.64 (d,
J ) 4.7 Hz, 3H), 1.34 (br s, 1H), 1.01 (d, J ) 6.7 Hz, 3H) ppm;
13C NMR (125 MHz, CDCl3) δ 136.3, 133.2, 129.7, 122.8, 60.7,
35.4, 35.3, 23.5, 21.7, 17.9 ppm; IR (film) 3325, 2963, 2870, 1449,
1377, 1042, 968, 868 cm-1; [R]D24 ) -59.3° (c ) 1.00, CHCl3);
HRMS (EI) exact mass calcd for C10H18O [M]+ 154.1358, found
154.1358.
Experimental Section
Preparation of Allylic Alcohol 7.30 To a solution of pentenyl
boronate 35 (124.7 mg, 0.410 mmol, 1 equiv) in hexanes (0.82
mL) at 0 °C was added methacrolein (68 µL, 0.82 mmol, 2 equiv).
The flask was capped, sealed with Teflon tape, and stirred in a
4 °C cold room for 36 h. The solution was warmed to rt, and the
hexanes were removed in vacuo (g50 mmHg). The residue was
dissolved in THF (1.92 mL), 3 M NaOH (0.48 mL) and 30% aq
H2O2 (0.38 mL) were added sequentially, and the mixture was
heated to reflux (bath temp 80 °C) for 2 h. The mixture was cooled
to rt and was extracted with Et2O (3 × 5 mL). The combined
organic layers were dried over Na2SO4, filtered, and concentrated
in vacuo (g50 mmHg). The product was purified via automated
silica column chromatography (0 f 20% Et2O/pentane over 20
min, 10 g column; TLC Rf ) 0.30, 20% Et2O/hexanes, anisaldehyde
stain) to provide allylic alcohol 7 (46.3 mg, 81%) as a clear,
colorless oil. Chiral GC (Chiraldex G-TA column, 30 m × 0.25
mm, 30 psi of He, 50 °C, 1 °C/min ramp) indicated the ee to be
98% (tmajor ) 24.38 min, tminor ) 25.62 min). Other data: 1H NMR
(500 MHz, CDCl3) δ 5.51 (dq, J ) 15.4, 6.4 Hz, 1H), 5.40 (ddq,
J ) 15.3, 7.1, 1.2 Hz, 1H), 4.95 (s, 1H), 4.88 (s, 1H), 3.90 (dd, J
) 4.5, 4.4 Hz, 1H), 2.36 (dq, J ) 12.8, 6.7 Hz, 1H), 1.71 (s, 3H),
1.68 (d, J ) 6.2 Hz, 3H), 1.56 (d, J ) 3.7 Hz, 1H), 0.98 (d, J )
6.9 Hz, 3H) ppm; 13C NMR (125 MHz, CDCl3) δ 145.9, 133.6,
125.3, 111.6, 78.8, 39.7, 18.6, 18.1, 14.2 ppm; IR (film) 3403, 2970,
2920, 2877, 1652, 1452, 1376, 1295, 1115, 1023, 981, 898, 557
Preparation of Thiazolidinethione Methyl Ether 47b. To a
100 mL flask was added (R)-N-acetyl-4-[CMe2(OTES)]-
thiazolidinethione (44b) (103.1 mg, 0.309 mmol, 2 equiv) in CH2-
Cl2 (1.5 mL). After being flushed with Ar, the stirring solution
was cooled to 0 °C, and PhBCl2 (0.041 mL, 0.309 mmol, 2 equiv)
was added dropwise via syringe. After 10 min, sparteine (0.142
mL, 0.618 mmol, 4 equiv) was added via syringe. The resulting
solution was warmed to rt and stirred for 30 min. The solution
was cooled to -78 °C, and bisoxazole dimethyl acetal 24b (52.7
mg, 0.155 mmol, 1 equiv) was added dropwise via cannula in CH2-
Cl2 (0.5 mL) followed by the dropwise addition of BF3‚OEt2 (0.030
mL, 0.232 mmol, 1.5 equiv). The solution was stirred at -78 °C
for 1.5 h and then was warmed to rt over 2 h. After being stirred
at rt for 1 h, the reaction was quenched with satd aq NH4Cl (5
mL). The contents of the flask were transferred to a separatory
funnel containing a 4:1 mixture of hexanes/CH2Cl2 (80 mL). The
layers were separated, the aqueous layer was re-extracted with CH2-
Cl2 (2 × 30 mL), and the combined organic layers were washed
with brine (1 × 20 mL), dried over anhydrous Na2SO4, filtered,
and concentrated in vacuo to give a yellow oil. 1H NMR integration
of the unpurified product indicated an 86:14 diastereomer ratio.
This material was purified via automated silica column chroma-
tography (0.7% f 3.5% acetone/CH2Cl2, 110 g column; TLC Rf
) 0.40 in 2% acetone/CH2Cl2) to provide diastereomerically pure
methyl ether 47b (53 mg, 53% yield) as a clear, yellow oil. In
(42) (a) Brooks, D. W.; Lu, L. D.-L.; Masamune, S. Angew. Chem., Int.
Ed. 1979, 18, 72. (b) Ohta, S.; Tsujimura, A.; Okamoto, M. Chem. Pharm.
Bull. 1981, 29, 2762-2768.
(43) Luche, J. L.; Gemal, A. L. J. Am. Chem. Soc. 1979, 101, 5848-
5849.
(44) (a) Seyferth, D.; Andrews, S. B. J. Organomet. Chem. 1971, 30,
151-166. (b) Still, W. C. J. Am. Chem. Soc. 1978, 100, 1481-1487.
J. Org. Chem, Vol. 73, No. 1, 2008 149

Smith et al.
much larger reactions where 24b was used in 0.712 and 1.807 g
amounts, the isolated yields were 48% and 46%, respectively. Other
data: 1H NMR (500 MHz, CDCl3) δ 7.65 (s, 1H), 5.28 (d, J ) 8.2
Hz 1H), 4.84 (dd, J ) 7.5, 5.1 Hz, 1H), 4.02 (dd, J ) 17.8, 7.6
Hz, 1H), 3.76 (dd, J ) 17.8, 5.2 Hz, 1H), 3.46 (dd, J ) 11.3, 8.1
Hz, 1H), 3.39 (d, J ) 11.4 Hz, 1H), 3.36 (s, 3H), 2.53 (s, 3H),
1.30 (s, 6H), 0.95 (t, J ) 7.5 Hz, 9H), 0.95 (s, 9H), 0.62 (q, J )
7.8 Hz, 6H), 0.381 (s, 3H), 0.377 (s, 3H) ppm; 13C NMR (125
MHz, CDCl3) δ 204.8, 170.5, 165.0, 156.6, 154.9, 141.1, 139.0,
135.4, 76.7, 72.5, 72.4, 57.0, 43.4, 29.9, 28.1, 26.5, 26.1, 17.6, 13.8,
7.0, 6.6, -5.9 ppm; IR (film) 2954, 2929, 2876, 1699, 1612, 1586,
1464, 1367, 1317, 1250, 1154 cm-1; [R]D25 ) -93.3° (c ) 1.02,
CHCl3); HRMS (CI) exact mass calcd for C29H50N3O5S2Si2 [M +
H]+ 640.2730, found 640.2745.
2H), 2.71 (dd, J ) 14.3, 6.1 Hz, 1H), 2.66 (dd, J ) 14.3, 6.1 Hz,
1H), 2.51 (dt, J ) 7.2, 6.7 Hz, 2H), 2.18-2.07 (m, 2H), 1.95 (ddd,
1H, J ) 12.8, 4.7, 1.5 Hz), 1.97 (ddt, J ) 12.3, 4.7, 2.4 Hz, 1H),
1.63 (d, J ) 4.0 Hz, 3H), 1.60 (d, J ) 1.4 Hz, 3H), 1.34 (dd, J )
12.6, 11.1 Hz, 1H), 1.28 (s, 3H), 1.22 (ddd, J ) 11.8, 11.8, 11.8
Hz, 1H), 0.98 (d, J ) 6.7 Hz, 3H), 0.95 (s, 9H), 0.86 (s, 9H), 0.39
(s, 3H), 0.38 (s, 3H), 0.04 (s, 3H), 0.03 (s, 3H) ppm; 13C NMR
(125 MHz, CDCl3) δ 168.0, 156.6, 154.7, 141.3, 139.0, 136.1,
136.0, 132.0, 130.4, 129.5, 128.7, 122.5, 99.6, 72.7, 65.7, 65.1,
56.3, 47.6, 45.3, 41.0, 40.3, 35.3, 35.2, 30.0, 28.2, 26.5, 25.8, 23.7,
23.3, 21.4, 18.0, 17.9, 17.6, -4.6, -5.9 ppm; IR (film) 2929, 2857,
1612, 1463, 1376, 1252, 1192, 1086, 969, 838, 778, 621, 525, 506,
489 cm-1; [R]D25 ) -40.3° (c ) 1.01, CHCl3); HRMS (CI) exact
mass calcd for C41H70N2O6Si2Li [M + Li]+ 749.4933, found
749.4917.
Preparation of Pyranone 53. To a 10 mL flask containing
alcohol 51 (238.0 mg, 0.431 mmol, 1 equiv) in DMF (2.7 mL)
under an Ar atmosphere was added magnesium acetoacetate (195
mg, 0.861 mmol, 2 equiv) and imidazole (32.2 mg, 0.474 mmol,
1.1 equiv). After being stirred for 29 h, the mixture was cooled to
0 °C in an ice bath, diluted with THF (3 mL), quenched with 7.5
mL of 2 M HCl, and stirred overnight. The layers were separated,
and the aqueous layer was extracted with EtOAc (4 × 20 mL).
The combined organic phases were dried over anhydrous Na2SO4,
filtered, and concentrated in vacuo. This material was purified via
automated silica column chromatography (55% f 80% EtOAc/
hexanes, 110 g column; TLC Rf ) 0.40 in 80% EtOAc/hexanes)
to provide dihydropyranone 53 (121.5 mg, 65% yield) as a clear
colorless oil: 1H NMR (500 MHz, CDCl3) δ 7.65 (s, 1H), 5.30 (s,
1H), 4.51-4.43 (m, 2H), 3.33 (s, 3H), 2.55 (s, 3H), 2.58-2.40
(m, 3H), 2.25 (ddd, J ) 13.5, 6.0, 6.0 Hz, 1H), 1.97 (s, 3H), 0.95
(s, 9H), 0.38 (s, 6H) ppm; 13C NMR (125 MHz, CDCl3) δ 192.4,
174.0, 165.1, 156.8, 154.8, 141.1, 138.8, 135.4, 104.8, 76.2, 72.6,
56.6, 40.6, 39.1, 26.3, 20.9, 17.5, 13.7, -6.00 ppm; IR (film) 2954,
2929, 2859, 1670, 1612, 1465, 1399, 1334, 1249, 1109, 1033, 842,
824, 781, 680 cm-1; [R]D25 ) -36.7° (c ) 0.96, CHCl3); HRMS
(CI) exact mass calcd for C22H33N2O5Si [M + H]+ 433.2159, found
433.2171.
Preparation of Bis-TBS-Protected Hennoxazole A (55). A 0.5
M solution of LiNEt2 was prepared by addition of n-BuLi (1.00
mL of a 2.72 M solution in hexanes, 2.72 mmol) to a 25 mL
concentration flask containing diethylamine (0.310 mL, 3.00 mmol)
in THF (4.13 mL) at -78 °C under an atmosphere of argon. After
5 min, the flask was warmed to 0 °C. Mixed methyl acetal 4a (R
) TBS) (69.0 mg, 0.119 mmol, 1 equiv) in a 10 mL concentration
flask was diluted with THF (2.0 mL). The solution was cooled to
-78 °C under an atmosphere of argon, and the LiNEt2 prepared
above was added dropwise until a yellow color persisted (to remove
any adventitious acid sourcesabout 4 drops). After this zero point,
LiNEt2 (0.285 mL of a 0.5 M solution in THF, 0.143 mmol, 1.2
equiv) was added dropwise via a gastight syringe. The reaction
took on a bright orange-red color. After the reaction mixture was
stirred at -78 °C for 15 min, allylic bromide 5 (31.8 mg, 0.131
mmol, 1.1 equiv) was added dropwise via cannula in THF (1.65
mL), causing the reaction to become progressively lighter orange-
yellow. After 15 min, the reaction was quenched with satd aq
NaHCO3 (5 mL), diluted with EtOAc (5 mL) and warmed to rt.
The layers were separated, and the aqueous phase was further
extracted with EtOAc (3 × 5 mL). The combined organics were
dried over anhydrous Na2SO4, filtered, and concentrated in vacuo.
This material was purified via automated silica column chroma-
tography (0 f 20% EtOAc/hexanes, TLC Rf ) 0.62 in 30% EtOAc/
hexanes) to provide protected hennoxazole 55 (67.7 mg, 77% yield)
as a clear colorless oil along with unreacted starting material 4a
(4.6 mg, 7%) and doubly alkylated product (6.1 mg, 6%): 1H NMR
(500 MHz, CDCl3) δ 7.62 (s, 1H), 5.51-5.38 (m, 2H), 5.38-5.31
(m, 2H), 4.99 (d, J ) 9.2 Hz, 1H), 4.45 (dd, J ) 8.1, 6.0 Hz, 1H),
3.97 (dddd, J ) 10.8, 10.8, 4.7, 4.7 Hz, 1H), 3.49-3.42 (m, 1H),
3.28 (s, 3H), 3.04-2.95 (m, 1H), 3.00 (s, 3H), 2.92 (t, J ) 7.6 Hz,
Preparation of (-)-Hennoxazole A (1a). To a 25 mL flask
containing protected hennoxazole 55 (67.7 mg, 0.091 mmol, 1
equiv) in THF (6.0 mL) under Ar(g) was added TBAF (547 µL of
a 1.0 M solution in THF, 0.547 mmol, 6 equiv). The solution was
stirred for 24 h, after which the reaction was concentrated in vacuo
and purified via flash chromatography (0 f 100% EtOAc, TLC Rf
) 0.27 in 80% EtOAc/hexanes) to provide 1a (46.2 mg, 99% yield)
as a clear colorless oil: 1H NMR (500 MHz, CDCl3) δ 8.13 (s,
1H), 7.63 (s, 1H), 5.50-5.38 (m, 2H), 5.37-5.31 (m, 2H), 4.99
(d, 1H, J ) 9.2 Hz, 1H), 4.45 (t, J ) 7.1 Hz, 1H), 4.04 (dddd, J )
10.8, 10.8, 4.5, 4.5 Hz, 1H), 3.61-3.54 (m, 1H), 3.31 (s, 3H), 3.07
(s, 3H), 3.03-2.95 (m, 1H), 2.90 (t, J ) 7.6 Hz, 2H), 2.71 (dd, J
) 14.4, 6.5 Hz, 1H), 2.66 (dd, J ) 14.4, 6.0 Hz, 1H), 2.51 (dt, J
) 7.2, 6.7 Hz, 2H), 2.22-2.15 (m, 1H), 2.11-2.03 (m, 2H), 1.97
(ddt, J ) 12.3, 4.7, 2.2 Hz, 1H), 1.70 (br s, 1H), 1.63 (dd, J ) 4.7,
1.0 Hz, 3H), 1.61 (d, J ) 1.2 Hz, 3H), 1.32 (s, 3H), 1.31 (dd, J )
12.4, 11.1 Hz, 1H), 1.22 (ddd, J ) 11.6, 11.6, 11.6 Hz, 1H), 0.98
(d, J ) 6.9 Hz, 3H) ppm; 1H NMR (500 MHz, acetone-d6) δ 8.41
(s, 1H), 8.00 (s, 1H), 5.53 (dt, J ) 15.2, 6.6 Hz, 1H), 5.44 (dt, J )
15.3, 6.4 Hz, 1H), 5.39-5.29 (m, 2H), 4.95 (d, J ) 9.4 Hz, 1H),
4.46 (dd, J ) 8.1, 6.2 Hz, 1H), 3.92-3.84 (m, 1H), 3.67 (d, J )
5.0 Hz, 1H), 3.54-3.47 (m, 1H), 3.22 (s, 3H), 3.02 (s, 3H), 3.05-
2.97 (m, 1H), 2.89 (t, J ) 7.4 Hz, 2H), 2.74-2.64 (m, 2H), 2.50
(dt, J ) 7.2, 6.7 Hz, 2H), 2.11-2.00 (m, 2H), 1.97 (ddd, J ) 12.4,
4.7, 1.5 Hz, 1H), 1.88 (ddt, J ) 12.3, 4.5, 2.3 Hz, 1H), 1.59 (s,
3H), 1.58 (s, 3H), 1.24 (s, 3H), 1.22 (dd, J ) 12.3, 11.3 Hz, 1H),
1.10 (ddd, J ) 11.7, 11.7, 11.7 Hz, 1H), 0.95 (d, J ) 6.9 Hz, 3H)
ppm; 13C NMR (125 MHz, CDCl3) δ 165.6, 155.6, 141.8, 138.1,
136.1, 135.8, 132.0, 130.4, 130.3, 129.7, 128.4, 122.5, 99.5, 72.9,
65.8, 64.6, 56.5, 47.7, 44.9, 40.5, 40.3, 35.2, 29.9, 28.2, 23.6, 23.3,
17.9 ppm; 13C NMR (125 MHz, acetone-d6) δ 166.0, 156.3, 142.2,
139.5, 137.6, 136.9, 132.6, 131.3, 130.9, 130.0, 129.6, 122.8, 100.0,
73.1, 66.4, 64.2, 56.0, 47.7, 45.8, 41.5, 41.3, 35.8, 35.7, 30.2, 28.5,
23.9, 23.3, 21.6, 17.9 ppm; IR (film) 3401, 2958, 2929, 1632, 1579,
1449, 1375, 1229, 1189, 1108, 1048, 1025, 969, 917, 830, 774
cm-1; [R]D25 ) -46.8° (c ) 0.635, CHCl3); HRMS (CI) exact
mass calcd for C29H42N2O6Li [M + Li]+ 521.3203, found 521.3185.
Acknowledgment. This work is dedicated to the memory
of Professor J. Hodge Markgraf (Williams College), a beloved
teacher, mentor, and friend. Support was provided by grants
from the National Science Foundation (CHE-0237658), the
American Chemical Society Petroleum Research Fund (36453-
GB1), and Williams College. Pfizer Summer Undergraduate
Research Fellowships to W.-H.K., E.P.B., V.D.B., and J.L.R.
are also gratefully acknowledged.
Supporting Information Available: Experimental procedures
and characterization data for compounds 4a, 5, 21-23, 24a-b,
27-30, 38-42, 45a-b, 46a-b, 47a, 48a, 51, and 54 and 1H NMR
and 13C NMR spectra for all new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
JO7018015
150 J. Org. Chem., Vol. 73, No. 1, 2008