Enantiospecific synthesis of (+)-byssochlamic acid, a nonadride from the ascomycete Byssochlamys fulva

Article abstract of DOI:10.1021/ja001898v

A photoaddition-cycloreversion strategy applicable to enantiospecific synthesis of natural byssochlamic acid and its enantiomer was developed in which porcine liver esterase (PLE) catalyzed hydrolysis of 24 was used to desymmetrize this dimethyl ester to give (R)-25. Analogous PLE catalyzed hydrolysis of dimethyl ester 32 and bis(methylthio)methyl ester 48, while highly regioselective, gave racemic half acids 33 and 49, respectively. Stepwise coupling of (±)-49 and (R)-46, the latter derived from 25, afforded diolide 52, which upon irradiation gave exo,exo and exo,endo [2 + 2] photoadducts 53 and 54, respectively. The photoadducts underwent thermal cycloreversion to produce nine-membered bislactones 55 and 56. Conversion of these lactones via 1,5-cyclononadienes 57 and 58 to (+)-byssochlamic acid (3) was accompanied by acid-catalyzed epimerization of the n-propyl substituent. Stepwise coupling of (S)-60 (the enantiomer of 46) with (±)-49 led to diolide 63, which upon photoaddition-cycloreversion gave bis-γ-lactones 66 and 67. These 1,5-cyclononadienes were transformed into (-)-68, the enantiomer of natural byssochlamic acid.

Full text of DOI:10.1021/ja001898v

J. Am. Chem. Soc. 2000, 122, 8665-8671
8665
Enantiospecific Synthesis of (+)-Byssochlamic Acid, a Nonadride
from the Ascomycete Byssochlamys fulVa
James D. White,* Jungchul Kim, and Nicholas E. Drapela
Contribution from the Department of Chemistry, Oregon State UniVersity, CorVallis, Oregon 97331-4003
ReceiVed May 30, 2000
Abstract: A photoaddition-cycloreversion strategy applicable to enantiospecific synthesis of natural bys-
sochlamic acid and its enantiomer was developed in which porcine liver esterase (PLE) catalyzed hydrolysis
of 24 was used to desymmetrize this dimethyl ester to give (R)-25. Analogous PLE catalyzed hydrolysis of
dimethyl ester 32 and bis(methylthio)methyl ester 48, while highly regioselective, gave racemic half acids 33
and 49, respectively. Stepwise coupling of (()-49 and (R)-46, the latter derived from 25, afforded diolide 52,
which upon irradiation gave exo,exo and exo,endo [2 + 2] photoadducts 53 and 54, respectively. The
photoadducts underwent thermal cycloreversion to produce nine-membered bislactones 55 and 56. Conversion
of these lactones via 1,5-cyclononadienes 57 and 58 to (+)-byssochlamic acid (3) was accompanied by acid-
catalyzed epimerization of the n-propyl substituent. Stepwise coupling of (S)-60 (the enantiomer of 46) with
(
1
()-49 led to diolide 63, which upon photoaddition-cycloreversion gave bis-γ-lactones 66 and 67. These
,5-cyclononadienes were transformed into (-)-68, the enantiomer of natural byssochlamic acid.
Introduction
The natural products known collectively as nonadrides
comprise a small structural class in which the core unit is a
1
nine-membered carbocyclic ring. Two five-membered anhy-
drides or an anhydride and a lactol are fused to the core, which
also bears a pair of n-alkyl chains and, in some cases, one or
more hydroxyl substituents. Glaucanic and glauconic acids, 1
and 2, respectively, were the first members of the class to be
discovered,2 and soon thereafter a “symmetrical” variant,
byssochlamic acid (3), was isolated by Raistrick from the
3
ascomycete Byssochlamys fulVa. Subsequently, two hepatotoxic
family of structures are CP-225,917 (9) and CP-263,114 (10),
two metabolites isolated by a research group at Pfizer from an
8
unidentified fungus which also produces zaragozic acid. The
substances, rubratoxins A (4) and B (5), were obtained from
4
extracts of the fungus Penicillium rubrum, and more recently
5
6
the nonadrides scytalidin (6), heveadride (7), and castaneiolide
(
7
8) have been found in nature. The latest examples of this
(
(
(
(
1) Sutherland, J. K. Fortschr. Chem. Org. Naturst. 1967, 25, 131.
2) Wijkman, N. Liebigs Ann. Chem. 1931, 485, 61.
3) Raistrick, H.; Smith, G. Biochem. J. 1933, 27, 1814.
4) (a) Wilson, B. J.; Wilson, C. H. J. Bacteriol. 1962, 83, 693. (b) B u¨ chi,
G.; Snader, K. M.; White, J. D.; Gougoutas, J. Z.; Singh, S. J. Am. Chem.
Soc. 1970, 92, 6638.
(
5) Overeem, J. C.; Mackor, A. Recl. TraV. Chim. Pays-Bas 1973, 92,
49.
6) Crane, R. I.; Hedden, P.; MacMillan, J.; Turner, W. B. J. Chem.
Soc., Perkin Trans. 1 1973, 194.
7) Arai, K.; Shimizu, S.; Miyajima, H.; Yamamoto, Y. Chem. Pharm.
Bull. 1989, 37, 2870.
9
3
powerful inhibition of ras farnesyl transferase by 9 and 10 has
(
(8) Dabrah, T. T.; Kaneko, T.; Massefski, W. Jr.; Whipple, E. B. J. Am.
Chem. Soc. 1997, 119, 1594.
(
(9) Leonard, D. M. J. Med. Chem. 1997, 40, 2971.
1
0.1021/ja001898v CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/18/2000

8666 J. Am. Chem. Soc., Vol. 122, No. 36, 2000
White et al.
made these nonadrides the objects of much interest,10 and a
synthesis of (()-9 and (()-10 was completed in 1999 by
Nicolaou.1 A unifying biogenetic hypothesis has been proposed
which accommodates the nonadrides, including 9 and 10, within
a matrix of head-to-head and head-to-tail dimers of a putative
dialkylmaleic anhydride intermediate.12
forded 13 in high yield, a result which can be readily understood
if 13 adopts a U-shaped conformation similar to that seen in
the X-ray crystal structure of the bis-p-bromophenylhydrazide
of byssochlamic acid. Before beginning our own synthesis of
3, the anticipated preference for the cis configuration of side
chains was substantiated through a conformational analysis in
which energy minimization using a PM3 algorithm predicted a
difference of ∼2.6 kcal/mol between cis (3c) and trans (3t)
isomers in favor of the former. The factor which destabilizes
1
Structural elucidation of 1, 2, and 3 by the Barton group13 at
Imperial College built upon earlier degradative work addressed
at 1 and 2 by Sutter14 and Kraft and led to the proposal 3,
exclusive of stereochemistry, for byssochlamic acid. Confirma-
tion of this assignment and designation of the cis relative
configuration was obtained through X-ray crystallographic
analysis of the bis-p-bromophenylhydrazide of 3.16 The absolute
configuration of byssochlamic acid was determined by degra-
dative experiments which caused fission of the nine-membered
ring and gave products of known stereochemistry.17 Unfortu-
nately, these degradative studies gave no hint of the relative
stability of cis and trans isomers of 3, a detail which proved to
be significant in our synthetic endeavors.
15
3
t relative to 3c is the pseudoaxial orientation of the propyl
chain, which creates a transannular steric interaction with an
endo hydrogen of the methylene adjacent to the ethyl substituent
(Figure 1). Interestingly, the nine-membered rings of both 3c
and 3t adopt a chairlike conformation, according to this
computation, and are, therefore, quite different in shape from
the ring conformation seen in the crystal structure of byssoch-
The first synthesis of a member of the nonadride family was
18
that of (()-3 by Stork. This pioneering accomplishment, which
created the nine-membered ring of 3 through Beckmann
fragmentation of oxime 11, provided the initial indication that
a cis orientation of ethyl and n-propyl substituents was more
stable. Reduction of 12 under thermodynamic conditions af-
1
6
lamic acid bis-p-bromophenylhydrazide. An important con-
sequence of this conformational analysis of 3 is that, if the center
bearing the propyl group is stereomutable, the absolute sense
of an asymmetric synthesis of byssochlamic acid can be
controlled through correct orientation of the remote ethyl
substituent. This point is addressed in more detail below.
A seemingly inconsequential addendum to the Imperial
College structural elucidation of byssochlamic acid17 was the
disclosure that the natural product undergoes a reaction to give
a saturated isomer when irradiated in tetrahydrofuran. Two
structures were considered for “photobyssochlamic acid,” one
(14) derived from intramolecular, parallel [2 + 2] cycloaddition
and the other (15) corresponding to a crossed photoaddition.
Since pyrolysis of the photoisomer of byssochlamic acid failed
to regenerate 3, the conclusion was drawn that its structure was
15. The implication that 14 should have reverted to byssochlamic
acid upon thermolysis was a proposition which played an
important role in guiding our synthesis plan for 3.
The concept of a [2 + 2] photoaddition-cycloreversion
strategy for assembling carbocyclic structures has long been
recognized as a powerful paradigm in medium-ring synthesis.19
It was first exemplified in the context of natural product
(
10) (a) Davies, H. M. L.; Calvo, R.; Ahmed, G. Tetrahedron Lett. 1997,
8, 1737. (b) Sgarbi, P. W. M.; Clive, D. L. J. Chem. Commun. 1997, 2157.
c) Armstrong, A.; Critchley, T. J.; Mortlock, A. A. Synlett 1998, 552. (d)
3
(
20
21
synthesis by Lange and Wender in their approaches to
germacranolide sesquiterpenes, and others soon followed their
Kwon, O.; Su, D.-S.; Meng, D.; Deng, W.; D’Amico, D. C.; Danishefsky,
S. J. Angew. Chem., Int. Ed. Engl. 1998, 37, 1877, 1880. (e) Waizumi, N.;
Itoh, T.; Fukuyama, T. Tetrahedron Lett. 1998, 39, 6015. (f) Chen, C.;
Layton, M. E.; Shair, M. D. J. Am. Chem. Soc. 1998, 120, 10784. (g)
Frontier, A. J.; Danishefsky, S. J.; Koppel, G. A.; Meng, D. Tetrahedron
2
2
lead. Our own plan for constructing the 1,5-cyclononadiene
nucleus of 3 (Scheme 1) hinged upon connection of two
(17) Baldwin, J. E.; Barton, D. H. R.; Sutherland, J. K. J. Chem. Soc.
1965, 1787.
1
998, 54, 12721. (h) Bio, M. M.; Leighton, J. L. J. Am. Chem. Soc. 1999,
1
21, 890.
(
(18) Stork, G.; Tabak, J. M.; Blount, J. F. J. Am. Chem. Soc. 1972, 94,
4735.
(19) Schaumann, E.; Ketcham, R. Angew. Chem., Int. Ed. Engl. 1982,
21, 225.
(20) (a) Lange, G. L.; Huggins, M.-A.; Neidert, E. Tetrahedron Lett.
1976, 4409. (b) Lange, G. L.; McCarthy, F. C. Tetrahedron Lett. 1978,
4749.
(21) (a) Wender, P. A.; Lechleiter, J. C. J. Am. Chem. Soc. 1977, 99,
267. (b) Wender, P. A.; Hubbs, J. C. J. Org. Chem. 1980, 45, 365. (c)
Wender, P. A.; Letendre, L. J. J. Org. Chem. 1980, 45, 367.
(22) (a) Wilson, S. R.; Phillips, L. R.; Pelister, Y.; Huffman, J. C. J.
Am. Chem. Soc. 1979, 101, 7373. (b) Williams, J. R.; Callahan, J. F. J.
Chem. Soc., Chem. Commun. 1979, 405. (c) Williams, J. R.; Callahan, J.
F. J. Org. Chem. 1980, 45, 4475, 4479. (d) Randall, M. L.; Lo, P. C.-K.;
Bonitatebus, P. J. Jr.; Snapper, M. L. J. Am. Chem. Soc. 1999, 121, 4534
and ref 5 cited.
11) (a) Nicolaou, K. C.; Baran, P. S.; Zhong, Y.-L.; Choi, H.-S.; Yoon,
W. H.; He, Y.; Fong, K. C. Angew. Chem., Int. Ed. Engl. 1999, 38, 1699.
b) Nicolaou, K. C.; Baran, P. S.; Zhong, Y.-L.; Fong, K. C.; He, Y.; Yoon,
W. H.; Choi, H.-S. Angew. Chem., Int. Ed. Engl. 1999, 38, 1676.
12) Moss, M. O. In Microbial Toxins; Ciegler, A., Ed.; Academic
Press: New York & London, 1971, 6, 381.
13) Baldwin, J. E.; Barton, D. H. R.; Bloomer, J. L.; Jackman, L. M.;
Rodriguez-Hahn, L.; Sutherland, J. K. Experientia 1962, 18, 345.
14) (a) Sutter, H.; Wijkman, N. Lieb. Ann. Chem. 1933, 505, 248. (b)
(
(
(
(
Sutter, H.; Wijkman, N. Lieb. Ann. Chem. 1935, 519, 97. (c) Sutter, H.;
Rottmayr, F.; Porsch H. Lieb. Ann. Chem. 1936, 521, 189.
15) (a) Kraft, K.; Porsch, H. Lieb. Ann. Chem. 1937, 527, 168. (b) Kraft,
K. Lieb. Ann. Chem. 1937, 530, 20.
16) Hamor, T. A.; Paul, I. C.; Monteath-Robertson, J.; Sim, G. A.
Experientia 1962, 18, 352.
(
(

Enantiospecific Synthesis of (+)-Byssochlamic Acid
J. Am. Chem. Soc., Vol. 122, No. 36, 2000 8667
our strategy to accommodate synthesis of either enantiomer of
byssochlamic acid with only minor modification.
Results and Discussion
The cyclopentene component 17 required for coupling with
16 was prepared from 4-ethylcyclohexanone (21) via bromina-
tion of the derived â-keto ester 22 and subsequent base-mediated
Favorskii rearrangement-elimination of dibromo keto ester 23.24
Figure 1. Energy-minimized (PM3) conformations of byssochlamic
acid (3c) and its trans isomer (3t).
Scheme 1
Desymmetrization of dimethyl ester 24 with buffered (pH 7)
2
5
porcine liver esterase gave a half acid 25 (>99% ee) in
virtually quantitative yield. The absolute configuration of 25
was determined as (R) by X-ray crystallographic analysis of
the (S)-(-)-R-methylbenzylamine salt 28 of carboxylic acid 27,
prepared via tert-butyl ester 26.
The cyclobutene partner 16 envisioned for coupling to 17
was prepared by photoaddition of bromomaleic anhydride 29
to 1-pentene. The resultant mixture of bromocyclobutanes 30
was hydrolyzed to give the dicarboxylic acid 31 as a mixture
of stereoisomers, and esterification with diazomethane followed
by elimination of hydrogen bromide then afforded cyclobutene
3
2 as a single product. Treatment of this dimethyl ester with
buffered porcine liver esterase furnished a carboxylic acid which
X-ray crystallographic analysis of its cyclohexylammonium salt
showed to be 33. The nearly quantitative yield of 33 implied
that the esterase had failed to distinguish enantiomers in the
racemate 32 while effecting a completely regioselective ester
hydrolysis. Optical rotation values for 33 and the derived alcohol
photopartners, cyclobutene 16 and cyclopentene 17, to produce
a substrate 18 which upon irradiation would be expected to yield
19. Ideally, the structure of this pentacycle should permit a
defined stereochemical relationship between ethyl and propyl
substituents which could be translated into the desired cis
configuration in 20. The constraints applied by the braces linking
A to C and B to D in 18 should forestall an undesired
cycloreversion pathway from 19 leading to a divinylcyclopen-
tane rather than 20. The general validity of these ideas was
demonstrated in a preliminary report describing a synthesis of
racemic 3.23 We have now extended the photoaddition-
cycloreversion approach to embrace an asymmetric variant of
the plan shown in Scheme 1 which leads to natural (+)-
byssochlamic acid as well as its (-) enantiomer. A subtle
element of stereocontrol is revealed in this work which enables
3
4 were zero, confirming that both compounds were indeed
racemic.
The failure of porcine liver esterase to distinguish stereo-
isomers of 32 while effecting highly regioselective hydrolysis
of this diester is presumably a reflection of the fact that the
rate of hydrolysis for each enantiomer is essentially the same,
thus precluding kinetic resolution. Although we must conclude
that the esterase, while recognizing the location of the propyl
substituent in relation to the two esters, makes no distinction
(24) Cf.: Schorno, K. S.; Adolphen, G. H.; Eisenbraun, E. J. J. Org.
Chem. 1969, 34, 2801.
(
23) White, J. D.; Dillon, M. P.; Butlin, R. J. J. Am. Chem. Soc. 1992,
(25) Sabbioni, G.; Shea, M. L.; Jones, J. B. J. Chem. Soc., Chem.
Commun. 1984, 236.
1
14, 9673.

8668 J. Am. Chem. Soc., Vol. 122, No. 36, 2000
White et al.
acid 25 with a monoprotected diol derived from 34 so that a
second coupling to yield a diolide could be accomplished in
sequential fashion. For this purpose, 34 was protected as its
silyl ether 36 and the methyl ester was reduced to alcohol 37.
Unfortunately, although 25 could be esterified cleanly with 37,
no means could be found for cleavage of the resulting methyl
ester while avoiding scission of the allylic ester linkage between
cyclobutene and cyclopentene. This forced us to consider an
alternative coupling of 37 with tert-butyl ester 27, an esterifi-
cation which proceeded efficiently to yield 38. Removal of the
silyl ether from 38 followed by acidic cleavage of the tert-butyl
ester from 39 afforded hydroxy acid 40, which underwent
2
7
Yamaguchi lactonization to produce diolide 41. To our
disappointment, all attempts to effect intramolecular photoad-
dition of 41 by irradiation through Pyrex resulted only in
recovered starting material, and after measurement of the UV
spectrum of 41, which showed an extinction coefficient (∈) of
500 at 313 nm, it became clear that insufficient light absorption
by 41 was responsible for this lack of reactivity. Irradiation of
41 at shorter wavelengths resulted in significant destruction of
the diolide, indicating that the synthesis plan outlined in Scheme
1 would not be implemented with this substrate.
between the orientations of the propyl group relative to the plane
of the cyclobutene, this type of observation is not unique.
2
6
Fortunately, the racemic nature of 33 is of no consequence to
the synthesis, because it was already known from our previous
work23 that the propyl group can be epimerized in favor of the
natural configuration at a late stage of the route to 3.
Our previous route to racemic byssochlamic acid23 did not
require a specific orientation of cyclobutene and cyclopentene
partners in the coupling process which led to 18, and in that
case, the two subunits were linked in a single step. However,
to transmute the chirality of 25 into an asymmetric pathway to
Conformational analysis of 41 revealed that the two carbonyl
groups are able to twist out of coplanarity with the cyclopentene
to a degree which would disrupt conjugation, and it is this factor
which is believed to be responsible for the unexpectedly poor
light absorption properties associated with this diolide. On the
other hand, conformational analysis suggested that an alternative
diolide in which the pair of lactone carbonyls is attached to the
cyclobutene would be more likely to have enforced coplanarity
and, therefore, enhanced UV absorption. A revised scheme was
therefore considered in which the cyclobutene and cyclopentene
functionality was reversed. This required reduction of the
carboxyl groups of 25 to differentiated primary alcohols without
passing through a meso intermediate, a sequence which was
accomplished by first reducing the carboxylic acid of 25 via its
3
, it was necessary to connect the partners stepwise in a
predetermined manner. Our initial goal was to couple carboxylic
2
8
mixed anhydride to the primary alcohol 42. The latter was
protected as THP ether 43 before the methyl ester was reduced
to furnish 44. Protection of this alcohol as silyl ether 45 was
(27) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull.
Chem. Soc. Jpn. 1979, 52, 1989.
(28) Soai, K.; Yokoyama, S.; Mochida, Y. Synthesis 1987, 647.
(26) Stein, K. A.; Toogood, P. L. J. Org. Chem. 1995, 60, 8110.

Enantiospecific Synthesis of (+)-Byssochlamic Acid
J. Am. Chem. Soc., Vol. 122, No. 36, 2000 8669
followed by selective removal of the THP protection with
magnesium bromide in ether to give 46.29
bearing the propyl substituent, and a similar inseparable mixture
of stereoisomers was seen in 52. In contrast to 41, diolide 52
exhibited significant absorption in its UV spectrum above 300
nm (∈ ∼ 16 000 at 313 nm), supporting the rationale based
upon comparison of the conformations of the two diolides
and suggesting a more productive outcome for the irradiation
of 52.
In the event, irradiation of 52 (1:1 mixture) through Pyrex
glass yielded two stereoisomeric photoproducts in approximately
equal amount. One of these products, resulting from the cis
isomer of 52, is assigned the exo,exo structure 53, in which
both the ethyl and propyl substituents are directed away from
the interior of the cage. The second photoproduct, derived from
the trans isomer of 52, is believed to be exo,endo adduct 54, in
which the ethyl substituent rather than the propyl group occupies
the interior space. Puckering of the cyclopentane ring of 54
moves the ethyl group away from the congested cage interior,
whereas the alternative photoadduct in which the propyl group
is endo would create severe compression between the propyl
substituent and a methylene group of the cyclopentane ring.
It was apparent during the photolysis of 52 that, even under
the irradiation conditions, cycloreversion was taking place, and
the facility of this fragmentation was confirmed when the
mixture of 53 and 54 was heated in toluene. A quantitative yield
The selection of a suitable cyclobutene derivative for coupling
with 46 was made with the goal of acquiring a diolide precursor
in which the silyl ether from 46 and the cyclobutene ester would
be cleaved in a single step that would not perturb the connecting
ester linkage between cyclobutene and cyclopentene subunits.
This reasoning led us to prepare the bismethylthiomethyl ester
4
8 from dicarboxylic acid 47, itself obtained by exhaustive
saponification of 32, in the hope that porcine liver esterase
hydrolysis of 48 would parallel the esterase catalyzed reaction
3
0
of 32. Our expectation was fulfilled with isolation of the
desired monocarboxylic acid 49, although a small quantity of
regioisomer 50 was also produced in this reaction. As with 33,
carboxylic acid 49 was found to be racemic. Methylthiomethyl
ester 51, obtained by coupling 49 with 46, was exposed to HF
which, as expected, removed both the silyl and methylthiomethyl
groups without compromising the cyclopentenylmethyl ester.
of 55 and 56 resulted from this reaction, the former arising from
5
3 and the latter from 54. The steps from 55 and 56 to (+)-
byssochlamic acid entailed hydrolysis of the pair of γ-lactones,
oxidation of the resultant diol to tetracarboxylic acids 57 and
58, and final treatment with hydrochloric acid to effect dehydra-
tion to give the bisanhydride (+)-3. As expected, the last step
was also accompanied by epimerization of the propyl substituent
in 58, probably via a maleic-itaconic anhydride equilibrium,
since only byssochlamic acid (3) and none of its trans isomer
was produced in this sequence. The identity of our synthesized
The hydroxy acid generated in this reaction was not isolated
(
+)-3 with natural byssochlamic acid was confirmed by direct
but was subjected directly to Keck-Steglich lactonization con-
1
13
ditions31 to produce dilactone 52. The remote stereocenters in
comparison (IR, H and C NMR, mass spectra, mp, optical
rotation) of the two samples.
5
1 provide no basis for stereoselection, and none was observed
With the recognition that the cis relationship of ethyl and
propyl substituents of 3 can be controlled through equilibration,
the opportunity presented by the strategy outlined in Scheme 1
for synthesis of the nonnatural enantiomer of byssochlamic acid
could be implemented in straightforward fashion by reversing
the configuration of the ethyl group in the cyclopentene 17. In
in the coupling of racemic 49 with 46; diester 51 was produced
as an inseparable 1:1 mixture of stereoisomers at the center
(
29) Kim, S.; Park, J. H. Tetrahedron Lett. 1987, 28, 439.
(30) Ladner, W. E.; Whitesides, G. M. J. Am. Chem. Soc. 1984, 106,
7
250.
31) Boden, E. P.; Keck, G. E. J. Org. Chem. 1985, 50, 2394.
(

8670 J. Am. Chem. Soc., Vol. 122, No. 36, 2000
White et al.
principle, there are two options for accomplishing this. A
different esterase could perhaps be employed to effect desym-
metrization of 24 in the opposite sense to 25.32 Alternatively,
The same hydrolysis, oxidation, and acidification sequence used
with 55 and 56 was applied to this mixture to give (-)-68,
enantiomeric with natural byssochlamic acid.
2
5 can be adapted to our goal by stepwise coupling with 49 in
In conclusion, we have shown that a [2 + 2] photoaddition-
cycloreversion pathway can be employed for asymmetric
synthesis of both enantiomers of the nonadride byssochlamic
acid. The possibility of extending this approach to more complex
members of the nonadride family (e.g. 5) can be foreseen, and
efforts along these lines will be described in due course.
the reverse sequence. This approach only requires interchange
of the primary alcohol and silyl ether substituents in 46, a
transformation which was easily achieved by silylation of 42
and reduction of methyl ester 59. The alcohol 60, enantiomeric
Experimental Section
Esterase Hydrolysis of 24. To a rapidly stirred mixture of 24 (1.70
g, 8.0 mmol) in acetone (20 mL) and pH 7 phosphate buffer (180 mL,
sodium phosphate dibasic and potassium phosphate dibasic) at room
temperature was added porcine liver esterase (200 mg, 4005 units).
After the mixture was stirred for 3 h at room temperature, it was diluted
with brine and ethyl acetate and was acidified to pH 1 with 2 N HCl
at 0 °C. The resulting emulsion was filtered through a pad of Celite,
and the filtrate was extracted with ethyl acetate (3 × 100 mL). The
combined organic extracts were washed with saturated aqueous NaCl,
dried over anhydrous MgSO
4
, and concentrated under reduced pressure
23
to yield 1.55 g (98%) of 25: [R]D -7.8 (c 1.00, CHCl
3
); IR (neat)
400-2400 (br), 2952, 2921, 2664, 1735, 1730, 1650, 1458, 1350,
3
1
(
-
1 1
293 cm ; H NMR (300 MHz, CDCl ) δ 0.9 (3H, t, J ) 7 Hz), 1.42
3
2H, dq, J ) 7, 7 Hz), 2.18 (1H, m), 2.48-2.63 (2H, m), 2.95-3.12
(
2H, m), 3.9 (3H, s); 13C NMR (75 MHz, CDCl
3
) δ 12.1, 28.3, 36.2,
+
4
1.2, 42.3, 53.5, 137.8, 147.3, 163.8, 168.6; MS(CI) m/z 199 (M
+
H), 181, 167, 151, 137; HRMS (CI) m/z 199.0969 (calcd for
: 199.0970).
Esterase Hydrolysis of 48. To a rapidly stirred mixture of 48 (0.155
10 15 4
C H O
g, 0.51 mmol) in acetone (2 mL) and pH 7 phosphate buffer (10 mL,
sodium phosphate dibasic and potassium phosphate dibasic) at room
temperature was added porcine liver esterase (5 mg, 100 units). After
the mixture was stirred for 3 h at room temperature, it was diluted
with brine (10 mL) and ethyl acetate (10 mL), and the resulting
emulsion was filtered through a pad of Celite. The filtrate was extracted
with ethyl acetate (4 × 30 mL), and the combined organic extracts
2
were washed with saturated aqueous NaCl, dried over anhydrous Na -
with 46, was used to esterify 49, and the resultant diester 61
was deprotected to 62 and lactonized as before to give a 1:1
mixture of inseparable diolides 63. Irradiation of this stereo-
isomeric mixture gave exo,exo and exo,endo cage photoadducts
4
SO , and concentrated under reduced pressure to yield 0.11 g (90%)
1
of a 7:1 mixture of 49 and 50 (determined by H NMR analysis) as
racemates. Chromatography of the mixture on silica, using 50% ether
in hexane as eluent, afforded 0.076 g of pure 49 as a colorless oil: IR
(neat) 3300-2800 (br), 2959, 2930, 1738, 1673, 1631, 1421, 1349,
6
4 and 65 in equal quantity, the assumption again being made
-
1 1
that in 65 the ethyl substituent rather than the propyl chain is
more easily accommodated in an endo orientation. As with 53/
3
1266, 1210 cm ; H NMR (300 MHz, CDCl ) δ 0.95 (3H, t, J ) 7
Hz), 1.43 (3H, m), 1.89 (1H, m), 2.31 (3H, s), 2.37 (1H, dd, J ) 2, 16
Hz), 2.89 (1H, dd, J ) 4, 16 Hz), 3.03 (1H, m), 5.36 (1H, d, J ) 12
Hz), 5.39 (1H, d, J ) 12 Hz), 12.03 (1H, br s); 13C NMR (100 MHz,
5
4, thermal cycloreversion of 64 and 65 was a quantitative
reaction, resulting in a 1:1 mixture of dilactones 66 and 67.
CDCl
3
) δ 14.0, 15.7, 20.4, 33.7, 34.1, 39.7, 71.0, 144.9, 150.2, 160.3,
+
1
(
64.3; MS(CI) m/z 245 (M + H), 195, 183, 167, 139, 123, 93; HRMS
CI) m/z 245.0846 (calcd for C11
Irradiation and Thermolysis of 52. A photolysis apparatus,
equipped with a dry ice condenser and an argon inlet, was charged
with a solution of 52 (0.040 mg, 0.13 mmol) in CH Cl (100 mL). The
17 4
H O S: 245.0848).
2
2
solution was purged with argon for 1 h and was irradiated with a 450-W
Hanovia medium-pressure mercury lamp using a Pyrex filter for 6 h.
The solvent was evaporated under reduced pressure, and the residue
was chromatographed on silica, using 5% ethyl acetate in hexane and
7
0% ethyl acetate in hexane as eluent, to give 7 mg of 52 and 19 mg
of a mixture of 53, 54, 55, and 56. The mixture was taken up into
toluene (3 mL) and heated to reflux for 7 h. After removal of the solvent,
the residue was chromatographed on silica, using 50% ethyl acetate in
hexane as eluent, to give 17.5 mg (44%, 56% based on recovered 52)
of 55 and 56 as a colorless oil: IR (neat) 2958, 2930, 1750, 1653,
-
1 1
1
1
1
3
457, 1060, 1033 cm ; H NMR (300 MHz, CDCl ) δ 0.88 (m, 3H),
.03 (m, 3H), 1.10-1.40 (m, 2H), 1.41-1.72 (m, 4H), 1.72-2.00 (m,
H), 2.25-2.58 (m, 5H), 2.70 (m, 1H), 2.96-3.28 (m, 1H), 4.43-
(32) (a) Wang, Y.-F.; Chen, C.-S.; Girdaukas, G.; Sih, C. J. Am. Chem.
Soc. 1984, 106, 3695. (b) Laumen, K.; Schneider, M. P. J. Chem. Soc.,
Chem. Commun. 1986, 1248.

Enantiospecific Synthesis of (+)-Byssochlamic Acid
J. Am. Chem. Soc., Vol. 122, No. 36, 2000 8671
4
2
4
1
2
.75 (m, 4H); 13C NMR (75 MHz, CDCl
0.7, 21.1, 27.7, 27.8, 29.0, 29.8, 30.3, 31.2, 32.1, 33.2, 34.5, 35.3,
1.6, 42.2, 70.8, 71.5, 71.8, 72.3, 127.4, 128.1, 130.1, 131.6, 158.2,
58.8, 159.2, 159.8, 173.9, 174.3, 174.8, 174.9; MS(CI) m/z 304 (M ),
86, 257, 224, 193, 167, 153, 143, 119, 99; HRMS (CI) m/z 304.1681
3
) δ 12.0, 12.3, 13.9, 14.0,
14 Hz), 2.77-2.98 (2H, m), 3.41 (1H, m); 13C NMR (75 MHz, CDCl
)
3
δ 11.6, 13.7, 20.6, 28.1, 29.2, 29.7, 30.0, 34.7, 36.0, 40.4, 143.2, 143.4,
+
144.1, 144.7, 164.9, 165.4 (2), 165.7; MS(CI) m/z 332 (M ), 260, 208,
+
20 6
166, 125; HRMS (CI) m/z 332.1263 (calcd for C18H O : 332.1259).
(
calcd for C18
+)-Byssochlamic Acid (3). To a solution of 55 and 56 (11 mg,
.036 mmol) in dioxane (2 mL) and water (2 mL) was added lithium
24 4
H O : 304.1675).
Acknowledgment. Dedicated to Professor Gilbert Stork. We
are indebted to Professor Alexandre F. T. Yokochi for the X-ray
crystal structures of 28 and 35, and to the late Professor Sir
Derek Barton for a sample of natural byssochlamic acid.
Financial support was provided by the National Science
Foundation (9711187-CHE).
(
0
hydroxide monohydrate (15 mg, 0.36 mmol), and the mixture was
stirred for 1.5 h at 50 °C. The mixture was cooled to 0 °C, and KMnO
40 mg, 0.252 mmol) was added. After 1 h at room temperature, the
solution was heated at 40 °C for 1 h, cooled to 0 °C, and acidified to
pH 1 with 2 N HCl. The mixture was extracted with CH Cl
(5 × 20
mL), and the combined organic extracts were washed with saturated
aqueous NaCl and dried over anhydrous MgSO . After removal of the
4
(
2
2
Supporting Information Available: Experimental proce-
dures for 3, 22, 24, 26, 27, 32-37, 39-56, 59-61, 63-68
4
1
13
solvent, the residue was chromatographed on silica, using 30% ethyl
acetate in hexane as eluent, to give 2.8 mg (23%) of (+)-3: mp 164-
1
1
7
1
(PDF); H and C NMR spectra of 3, 22, 24-27, 32-37, 39-
4
9, 51, 52, 55, 56, 59-61, 63, 66-68; X-ray crystallographic
data for 28 and 35. This material is available free of charge via
2
D
3
65 °C; [R] +101 (c 0.24, CHCl
3
); IR (neat) 2966, 2934, 1829,
-
1 1
766, 1260, 927 cm ; H NMR (300 MHz, CDCl ) δ 0.96 (3H, t, J )
3
the Internet at http://pubs.acs.org.
Hz), 1.12 (3H, t, J ) 7 Hz), 1.28-1.50 (2H, m), 1.50-1.75 (4H, m),
.90 (1H, m), 2.25-2.43 (2H, m), 2.65 (1H, m), 2.72 (1H, dd, J ) 2,
JA001898V