Enantioselective total synthesis of (-)-chlorothricolide via the tandem inter- and intramolecular Diels-Alder reaction of a hexaenoate intermediate

Article abstract of DOI:10.1021/ja980611f

An enantioselective total synthesis of (-)-chlorothricolide (1) has been completed via a route involving the tandem inter- and intramolecular Dieis-Alder (IMDA) reaction of hexaenoate 19 and the chiral dienophile (R)-12. This reaction, which establishes s

Full text of DOI:10.1021/ja980611f

J. Am. Chem. Soc. 1998, 120, 7411-7419
7411
Enantioselective Total Synthesis of (-)-Chlorothricolide via the
Tandem Inter- and Intramolecular Diels-Alder Reaction of a
Hexaenoate Intermediate
William R. Roush*,1 and Richard J. Sciotti2
Contribution from the Department of Chemistry, Indiana UniVersity, Bloomington, Indiana 47405, and
Department of Chemistry, UniVersity of Michigan, Ann Arbor, Michigan 48109
ReceiVed February 23, 1998
Abstract: An enantioselective total synthesis of (-)-chlorothricolide (1) has been completed via a route involving
the tandem inter- and intramolecular Diels-Alder (IMDA) reaction of hexaenoate 19 and the chiral dienophile
(R)-12. This reaction, which establishes seven asymmetric centers in a single operation, is feasible only by
virtue of the high diastereofacial and exo selectivity of dienophile 12. The C(9)-trimethylsilyl steric directing
group of 19 also plays a key role by controlling the stereochemical course of the IMDA reaction leading to the
bottom half octahydronaphthalene unit. Hexaenoate 19 was prepared in 32% overall yield by a 10-step sequence
starting from the known acetylenic ketone 33. Key steps include the asymmetric reduction of 33 using Alpine
Borane (up to 94% ee), the Suzuki cross coupling of R-iodo vinylsilane 20 with vinylboronic acid 21, and the
Horner-type olefination of aldehyde 41 with dienylic phosphonate 22. The key tandem inter-intramolecular
Diels-Alder reaction was performed by heating a mixture of 19 and (R)-12 (2 equiv) in toluene at 120 °C,
which provided the targeted double cycloadduct 44 in 40-45% yield, along with 19% of other cycloadduct
isomers and 25-20% of the IMDA adduct 24 with an (E,E,E)-C(16)-C(21) triene. The latter compound was
recycled by treatment with additional (R)-12 in trichloroethylene at 125 °C. The yield of 44 from hexaenoate
19 was 55-59% after one recycle of (E,E,E)-24. Elaboration of 44 to (-)-chlorothricolide was accomplished
by a 9-step sequence in 26% overall yield, key steps of which included the construction of the spirotetronate
subunit of 51 via the Dieckmann cyclization of 50, deprotection of the two allyl units with Pd(0) catalysis, and
the BOP-Cl-mediated macrolactonization of seco acid 52. The vinyl trimethylsilane substituent was removed
in the final step of the synthesis by treatment with EtSH and BF3‚Et2O. Because an authentic sample of
chlorothricolide was not available, synthetic (-)-chlorothricolide was treated with CH2N2 to give 24-O-methyl
chlorothricolide methyl ester (59) [[R]25D -29.3° (c ) 0.95, CHCl3), mp 228.5-229 °C; lit.1a [R]D20 -30° (c
) 1, CHCl3), lit.1a mp 230 °C)] which proved identical in all respects (except optical rotation and mp) with
an authentic sample of racemic 59 provided by Professor Yoshii.
Chlorothricolide (1) is the aglycon of chlorothricin that was
isolated from Streptomyces antibioticus by Keller-Schierlein and
co-workers in 1969.3,4 Chlorothricin is an inhibitor of pyruvate
carboxylase and is active against Gram-positive bacteria.5,6
Chlorothricolide methyl ester retains some antibiotic activity,
although at concentrations 4 to 10-fold higher than the effective
concentration of chlorothricin itself. The structure of chloro-
thricolide was originally assigned by using spectroscopic and
degradation procedures3 and, ultimately, was confirmed by
single-crystal X-ray analysis of the cesium salt of chlorothri-
colide methyl ester.4
Interest in chlorothricolide as a synthetic target stems from
the fact that it is the parent compound in a growing family of
natural products possessing spirotetronic acid units. Other
(1) Address correspondence to this author at the Department of Chem-
istry, University of Michigan, Ann Arbor, MI 48109-1055; e-mail:
roush@umich.edu.
members of this family include kijanimicin,7 tetrocarcin,8
pyrrolosporin A,9 PA-46101A and B,10 the gastric ATP-ase
(2) Taken in part from: Sciotti, R. J. Ph.D. Thesis, 1994, Indiana
University, Bloomington, IN 47405.
(3) Muntwyler, R.; Keller-Schierlein, W. HelV. Chim. Acta 1972, 55,
2071.
(4) Brufani, M.; Cerrini, S.; Fedeli, W.; Mazza, F.; Muntwyler, R. HelV.
Chim. Acta 1972, 55, 2094.
(5) Schindler, P. W.; Zahner, H. Eur. J. Biochem. 1973, 39, 591.
(6) Schindler, P. W. Eur. J. Biochem. 1975, 51, 579 and references
therein.
(7) Mallams, A. K.; Puar, M. S.; Rossman, R. R.; McPhail, A. T.;
Macfarlane, R. D.; Stephens, R. L. J. Chem. Soc., Perkin Trans. 1 1983,
1497.
(8) Hirayama, N.; Kasai, M.; Shirahata, K.; Ohashi, Y.; Sasada, Y. Bull.
Chem. Soc. Jpn. 1982, 55, 2984.
(9) Schroeder, D. R.; Colson, K. L.; Klohr, S. E.; Lee, M. S.; Matson,
J. A.; Brinen, L. S.; Clardy, J. J. Antibiot. 1996, 49, 865.
(10) Matsumoto, M.; Kawamura, Y.; Yoshimura, Y.; Terui, Y.; Nakai,
H.; Yoshida, T.; Shoji, J. J. Antibiot. 1990, 43, 739.
S0002-7863(98)00611-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/21/1998

7412 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Roush and Sciotti
Chart 1
of kijanimicin, and tetronolide (3), the aglycon of the
tetrocarcins.38-56 Total syntheses of tetronolide50 and 24-O-
methyl chlorothricolide (4)31 have been accomplished by Yoshii,
and a formal synthesis of tetronolide56 has been achieved in
our laboratory. We report herein the full details of our
enantioselective total synthesis of (-)-chlorothricolide, a pre-
liminary report of which appeared in 1994.36
Synthetic Analysis
At the outset of our work on chlorothricolide, we planned to
develop highly stereoselective syntheses of suitable top and
bottom half units, and then to couple these fragments at a late
stage of the synthesis. Toward this end, we established in 1988
that the intramolecular Diels-Alder (IMDA) reaction57-59 of 5
provided the trans-fused perhydronaphthalene intermediate 6
with good selectivity and demonstrated that 6 is easily elaborated
into the chlorothricolide bottom half fragment 9.26,34 The C(9)-
TMS substituent plays a critical role in controlling the stereo-
chemical course of this IMDA reaction, as the major product
of cyclizations of related substrates lacking the C(9)-TMS
substituent is a cis-fused cycloadduct analogous to 7.15 A full
analysis of the steric directing group strategy has been published
elsewhere.34
In 1992 we reported a highly enantio- and diastereoselective
synthesis of the top half spirotetronate fragment 1435,37 via the
highly exo, regio- and diastereofacially selective bimolecular
Diels-Alder reaction of trienoate 11 and the chiral dienophile
(R)-12.60 We have used similar sequences to synthesize the
spirotetronate units of kijanolide and tetronolide,46,52,53,56 in all
(32) Hirsenkorn, R.; Schmidt, R. R. Liebigs Ann. Chem. 1990, 9, 883.
(33) Hirsenkorn, R.; Haag-Zeino, B.; Schmidt, R. R. Tetrahedron Lett.
1990, 31, 4433.
(34) Roush, W. R.; Kageyama, M.; Riva, R.; Brown, B. B.; Warmus, J.
S.; Moriarty, K. J. J. Org. Chem. 1991, 56, 1192.
(35) Roush, W. R.; Sciotti, R. J. Tetrahedron Lett. 1992, 33, 4691.
(36) Roush, W. R.; Sciotti, R. J. J. Am. Chem. Soc. 1994, 116, 6457.
(37) Roush, W. R.; Sciotti, R. J. J. Org. Chem. In press.
(38) Takeda, K.; Kato, H.; Sasahara, H.; Yoshii, E. J. Chem. Soc., Chem.
Commun. 1986, 1197.
(39) Marshall, J. A.; Grote, J.; Shearer, B. J. Org. Chem. 1986, 51, 1633.
(40) Takeda, K.; Urahata, M.; Yoshii, E.; Takayanagi, H.; Ogura, H. J.
Org. Chem. 1986, 51, 4735.
inhibitors A88696 C, D, and F,11 and the quartromicins12 (Chart
1). Virtually all of the published synthetic work in this area
has focused on chlorothricolide,13-37 kijanolide (2), the aglycon
(11) Bonjouklian, R.; Mynderse, J. S.; Hunt, A. H.; Deeter, J. B.
Tetrahedron Lett. 1993, 34, 7857.
(12) Kusumi, T.; Ichikawa, A.; Kakisawa, H.; Tsunakawa, M.; Konishi,
M.; Oki, T. J. Am. Chem. Soc. 1991, 113, 8947.
(13) Ireland, R. E.; Thompson, W. J. J. Org. Chem. 1979, 44, 3041.
(14) Ireland, R. E.; Thompson, W. J.; Srouji, G. H.; Etter, R. J. Org.
Chem. 1981, 46, 4863.
(15) Roush, W. R.; Hall, S. E. J. Am. Chem. Soc. 1981, 103, 5200.
(16) Hall, S. E.; Roush, W. R. J. Org. Chem. 1982, 47, 4611.
(17) Snider, B. B.; Burbaum, B. W. J. Org. Chem. 1983, 48, 4370.
(18) Schmidt, R. R.; Hirsenkorn, R. Tetrahedron Lett. 1984, 25, 4357.
(19) Marshall, J. A.; Audia, J. E.; Grote, J. J. Org. Chem. 1984, 49,
5277.
(20) Boeckman, R. K., Jr.; Barta, T. E. J. Org. Chem. 1985, 50, 3421.
(21) Roush, W. R.; Kageyama, M. Tetrahedron Lett. 1985, 26, 4327.
(22) Ireland, R. E.; Varney, M. D. J. Org. Chem. 1986, 51, 635.
(23) Marshall, J. A.; Audia, J. E.; Shearer, B. G. J. Org. Chem. 1986,
51, 1730.
(41) Takeda, K.; Yano, S.; Sato, M.; Yoshii, E. J. Org. Chem. 1987, 52,
4135.
(42) Marshall, J. A.; Grote, J.; Audia, J. E. J. Am. Chem. Soc. 1987,
109, 1186.
(43) Takeda, K.; Kobayashi, T.; Saito, K.; Yoshii, E. J. Org. Chem. 1988,
53, 1092.
(44) Takeda, K.; Yano, S.-g.; Yoshii, E. Tetrahedron Lett. 1988, 29, 6951.
(45) Roush, W. R.; Brown, B. B.; Drozda, S. E. Tetrahedron. Lett. 1988,
29, 3541.
(46) Roush, W. R.; Brown, B. B. Tetrahedron Lett. 1989, 30, 7309.
(47) Marshall, J. A.; Salovich, J. M.; Shearer, B. G. J. Org. Chem. 1990,
55, 2398.
(48) Boeckman, R. K., Jr.; Barta, T. E.; Nelson, S. G. Tetrahedron Lett.
1991, 32, 4091.
(49) Boeckman, R. K., Jr.; Estep, K. G.; Nelson, S. G.; Walters, M. A.
Tetrahedron Lett. 1991, 32, 4095.
(50) Takeda, K.; Kawanishi, E.; Nakamura, H.; Yoshii, E. Tetrahedron
Lett. 1991, 32, 4925.
(24) Marshall, J. A.; Audia, J. E.; Grote, J.; Shearer, B. G. Tetrahedron
1986, 42, 2893.
(25) Marshall, J. A.; Shearer, B. G.; Crooks, S. L. J. Org. Chem. 1987,
52, 1236.
(26) Roush, W. R.; Riva, R. J. Org. Chem. 1988, 53, 710.
(27) Danishefsky, S. J.; Audia, J. E. Tetrahedron Lett. 1988, 29, 1371.
(28) Okumura, K.; Okazaki, K.; Takeda, K.; Yoshii, E. Tetrahedron Lett.
1989, 30, 2233.
(51) Marshall, J. A.; Xie, S. J. Org. Chem. 1992, 57, 2987.
(52) Roush, W. R.; Koyama, K. Tetrahedron Lett. 1992, 33, 6227.
(53) Roush, W. R.; Brown, B. B. J. Org. Chem. 1993, 58, 2151.
(54) Roush, W. R.; Brown, B. B. J. Am. Chem. Soc. 1993, 115, 2268.
(55) Boeckman, R. K., Jr.; Wrobleski, S. T. J. Org. Chem. 1996, 61,
7238.
(56) Roush, W. R.; Reilly, M. L.; Koyama, K.; Brown, B. B. J. Org.
Chem. 1997, 62, 8708.
(29) Poss, A. J.; Brodowski, M. H. Tetrahedron Lett. 1989, 30, 2505.
(30) Roth, G. P.; Rithner, C. D.; Meyers, A. I. Tetrahedron 1989, 45,
6949.
(31) Takeda, K.; Igarashi, Y.; Okazaki, K.; Yoshii, E.; Yamaguchi, K.
J. Org. Chem. 1990, 55, 3431.
(57) Ciganek, B. Org. React. 1984, 32, 1.
(58) Craig, D. Chem. Soc. ReV. 1987, 16, 187.
(59) Roush, W. R. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Ed.; Pergamon Press: Oxford, 1991; Vol. 5, pp 513-550.
(60) Roush, W. R.; Brown, B. B. J. Org. Chem. 1992, 57, 3380.

Total Synthesis of (-)-Chlorothricolide
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7413
ate, which we imagined could be assembled in a straightforward
manner from vinyl iodide 20, vinylboronic acid 21, dienylic
phosphonate 22, and â-keto phosphonate 23.
cases taking full advantage of the remarkable exo and diaste-
reofacial selectivity of (R)-12.60
Having developed selective and efficient routes to 9 and 14,
we turned our attention to the coupling of these intermediates
via formation of the C(14)-C(15) bond. It was readily apparent
that this step would present a significant challenge.22 If the
bottom half fragment was used as the nucleophilic component
(e.g., 15, W ) -SO2Ar) in an alkylation reaction with a top
half electrophile 16 (X ) Br, I, etc.), it would be necessary to
develop conditions to minimize addition of the C(14) carbanion
to the C(1) carboxyl group (perhaps by using a carboxylate in
the coupling sequence). While use of the top half fragment as
the nucleophilic component (18, W ) SO2Ar) in an alkylation
reaction with an electrophilic bottom half (17, X ) leaving
group) might pose fewer problems, we nevertheless were
apprehensive about the prospects of performing this coupling
on such highly functionalized intermediates.
However, use of 19 as a key synthetic intermediate raises
questions about selectivity of the proposed tandem inter- and
intramolecular Diels-Alder reaction. Three different modes
of intramolecular Diels-Alder reactions are possible with this
system: (i) addition of the C(2)-C(3) dienophile across the
C(8)-C(11) diene, leading to 24; (ii) addition of the C(16)-
C(17) olefin across the C(8)-C(11) diene, leading to 25; and
(iii) participation of the C(10)-C(11) olefin as a dienophile in
a IMDA reactions with the C(16)-C(19) diene, leading to 26.
There are also three different dienes that, in principle, can
undergo bimolecular Diels-Alder reactions with dienophile (R)-
As an alternative, we recognized that coupling of precursors
to the top and bottom half fragmentssprior to the two Diels-
Alder reactionsswould pose relatively few complications.
Hexaenoate 19 was thus identified as a key synthetic intermedi-

7414 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Roush and Sciotti
to be 89-94% ee by Mosher ester analysis.67 Protection of
the secondary hydroxyl group as a MOM ether provided 34 in
85% yield for the two steps. This sequence introduces the only
stereocenter of chlorothricolide that is not established in the
tandem intra-intermolecular Diels-Alder reaction. Partial
reduction of the carbomethoxyl unit of 34 with 1.05 equiv of
DIBAL-H in CH2Cl2 at -78 °C and subsequent protection of
the aldehyde as a dimethyl acetal afforded 35. Hydroalumi-
nation of 35 using DIBAL-H in Et2O at 50 °C in a sealed tube,
followed by iodination of the resulting vinylalane intermediate
provided vinyl iodide 20 in 82% yield.68,69
12.61 Taking into account all possible endo, exo, diastereofacial
(with respect to 12), and regiochemical possibilities afforded
by these diene-dienophile combinations, there are 96 distinct
double Diels-Alder adducts that could be produced from 19
and (R)-12 (assuming both components are enantiomerically
pure).
Fortunately, knowledge of the relative rates of the various
Diels-Alder reactions enabled us to rule out all but the desired
pathway as reasonable possibilities. Of the three potential
modes of intramolecular Diels-Alder reactions, the C(2)-C(3)
dienophile/C(8)-C(11) diene combination leading to 24 was
expected to be the fastest since the C(2)-C(3) dienophile is
the most activated of the three dienophiles.57-59 Concerning
the three bimolecular Diels-Alder combinations, previous
studies in our laboratory indicated that the rates of Diels-Alder
reactions of dienophiles (R)-12 and 28 with acyclic dienes such
as 2756,62 are considerably slower than their reactions with
conjugated trienes such as 30.35,37 On this basis, we regarded
the C(8)-C(11) diene unit of 19 as an unlikely reaction partner
with (R)-12. In addition, the fact that the Diels-Alder reaction
of 30 and (R)-12 proceeds with exceptionally high exo, regio-
and diastereofacial selectivity gave us confidence that the
addition of (R)-12 across the C(18)-C(21) unit of 19 would be
comparably selective.35,37
Results and Discussion
Synthesis of Vinyl Iodide 20. The synthesis of vinyl iodide
20 was initiated by acylation of bis(trimethylsilyl)acetylene with
commercially available acid chloride 32, which provided the
known acetylenic ketone 33 in 70% yield.63 Asymmetric
reduction of 33 with Alpine Borane,64-66 generated in situ by
heating a neat mixture of (-)-R-pinene and 9-BBN, afforded
the optically active alcohol in 82% yield.63 The enantiomeric
purity of this intermediate from various runs was determined
Synthesis of Dienylic Phosphonate 22. Phosphonate 22 was
initially prepared from enal 3653 by way of the known dienylic
alcohol 37.53 However, all attempts to convert 37 to the
corresponding allylic bromide 38 provided a ca. 3:1 mixture of
38 and the isomeric dienylic bromide 39 that could not be
separated.70 Subjection of this mixture to an Arbuzov reac-
tion71,72 with triethyl phosphite gave phosphonate 22 in 60%
overall yield, with no evidence that an allylic phosphonate
(61) Fringuelli, F.; Taticchi, A. Dienes in the Diels-Alder Reaction;
Wiley: New York, 1990.
(62) Roush, W. R.; Essenfeld, A. P.; Warmus, J. S.; Brown, B. B.
Tetrahedron Lett. 1989, 30, 7305.
(63) Chemin, D.; Linstrumelle, G. Tetrahedron 1992, 48, 1943.
(64) Midland, M. M.; McDowell, D. C.; Hatch, R. L.; Tramontano, A.
J. Am. Chem. Soc. 1980, 102, 867.
(65) Brown, H. C.; Pai, G. G. J. Org. Chem. 1985, 50, 1384.
(66) Brown, H. C.; Ramachandran, P. V. Acc. Chem. Res. 1992, 25, 16.
(67) Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34,
2543.
(68) Hasan, I.; Kishi, Y. Tetrahedron Lett. 1980, 21, 4229.
(69) On, H. P.; Lewis, W.; Zweifel, G. Synthesis 1981, 999.
(70) Bringmann, G.; Schneider, S. Synthesis 1983, 139.
(71) Arbuzov, B. A. Pure Appl. Chem. 1964, 9, 307.
(72) Bhattacharya, A. K.; Thyagarajan, G. Chem. ReV. 1981, 81, 415.

Total Synthesis of (-)-Chlorothricolide
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7415
Hydrolysis of the dimethyl acetal unit of 42 was surprisingly
difficult. Use of standard hydrolysis conditions (oxalic acid,
acetone; PPTs, wet acetone; HOAc, acetone; trifluoroacetic acid,
acetone; trifluoroacetic acid, CHCl3; or p-TsOH, acetone)
resulted in recovery of 42 or considerable decomposition of 42
without product formation. Fortunately, deprotection of the
deriving from a SN2 substitution of 39 was produced. This result
suggested that dienylic alcohol 37 is not an obligatory inter-
mediate and implied that 22 could be prepared more directly
by way of allylic alcohol 40. Indeed, bromination of 40
provided a mixture of 38 and 39 that was very comparable in
composition to the mixture prepared from 37. Arbuzov reaction
of this mixture with triethyl phosphite then provided 22 in 55-
60% overall yield from 40.
dimethyl acetal without decomposition of the pentaene could
be accomplished by using LiBF4 in wet CH3CN.79 Finally, the
C(1)-C(3) dienophile unit was introduced via Horner-Wads-
worth-Emmons olefination of the resulting aldehyde with
â-keto phosphonate 23, thereby providing hexaenoate 19 in 78%
yield.80 â-Keto phosphonate 23 was prepared in 88% overall
yield by selective hydrolysis of the commercially available
phosphonopropionate 43, followed by DCC coupling81 of the
carboxylic acid with allyl alcohol. Overall, hexaenoate 19 was
assembled by a highly convergent, 10-step synthesis that
proceeds in 32% yield from the known acetylenic ketone 33.
Phosphonate 22 prepared by this route was obtained as a ca.
5:1 mixture of C(20)-trisubstituted olefin isomers. This stereo-
chemical imperfection was not viewed as a serious problem,
since in previous studies we noted that this trisubstituted olefin
isomerizes (reversibly) under the conditions of the Diels-Alder
reaction with (R)-12 and that only the (Z)-C(20) isomer
undergoes the Diels-Alder reaction.35,37
Synthesis of Hexaenoate 19. The assembly of hexaenoate
19 began with the Suzuki cross-coupling73 of vinyl iodide 20
and vinylboronic acid 2137 using Kishi’s modified conditions.74
Swern oxidation of the resulting primary alcohol then provided
aldehyde 41 in 86% yield for the two steps.75,76 The C(16)-
C(21) triene unit was elaborated by olefination of aldehyde 41
with the lithium anion of dienylic phosphonate 22 (a ca. 5:1
mixture of C(20) (Z)- and (E)-olefin isomers).77,78 Pentaene
42 was obtained in 84% yield, also as a 5:1 mixture of C(20)
(Z)- and (E)-olefin isomers. However, only the (E)-isomer of
the newly formed C(16)-C(17) olefin was detected.
The Tandem Intra-Intermolecular Diels-Alder Reaction.
The key tandem intra-intermolecular Diels-Alder reaction was
performed by heating a 1 M solution of hexaenoate 19 and 2
equiv of dienophile (R)-1260 in toluene at 120 °C for 20 h in
the presence of a crystal of BHT as a radical inhibitor. This
provided the desired adduct 44 in 40-45% yield as well as
19% of a mixture of other cycloadducts. In addition, 25-30%
(73) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457.
(74) Uenishi, J.-i.; Beau, J.-M.; Armstrong, R. W.; Kishi, Y. J. Am. Chem.
Soc. 1987, 109, 4756.
(75) Mancusco, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43,
2480.
(79) Lipshutz, B. H.; Harvey, D. F. Synth. Commun. 1982, 12, 267.
(80) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(81) Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17,
522.
(76) Tidwell, T. T. Synthesis 1990, 857.
(77) Maryanoff, B. E.; Reitz, A. B. Chem. ReV. 1989, 89, 863.
(78) Thomas, E. J.; Whitehead, J. W. F. J. Chem. Soc., Perkin Trans. 1
1989, 499.

7416 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Roush and Sciotti
of the intramolecular Diels-Alder adduct (E,E,E)-24 with an
isomerized C(20)-C(21)-trisubstituted double bond was also
obtained. Because we had already noticed that the C(20)-
C(21)-trisubstituted olefin readily isomerizes under the condi-
tions employed for this Diels-Alder reaction,35,37 (E,E,E)-24
was treated with additional dienophile (R)-12 (2.0 equiv) in
trichloroethylene (1.5 M) at 125 °C for 28-36 h in the presence
of BHT as a radical inhibitor. This provided cycloadduct 44
in up to 58% yield, along with additional (E,E,E)-24 (20-45%)
that could be further recycled. The yield of the desired double
Diels-Alder adduct 44 was 55-59% from 19 after one such
recycle.
of (R)-12 and 30,35,37 and that the intramolecular Diels-Alder
reaction of the C(1)-C(11) undecatrienoate would proceed with
72:19:9 selectivity, as modeled by the IMDA cyclization of 5.34
That is, on the basis of these previously studied examples, one
would anticipate that 44 should be the major product of a
67:18:8:5:1:1 mixture of double Diels-Alder adducts. In fact,
HPLC analysis of the crude Diels-Alder reaction mixture
indicated that four predominant double cycloadducts were
obtained in the ratio of 67:13:5:5 (average of several runs). The
HPLC trace also contained several other minor bands that
presumably represent minor diastereomeric products expected
from the Diels-Alder reaction of (R)-12 and ent-19 (present at
the 3-5% level in 19).
HPLC separation of the mixed fractions obtained from the
initial double Diels-Alder reaction provided 45, corresponding
to the second most abundant double cycloadduct in the HPLC
trace of the crude reaction mixture. This compound was
assigned a cis-fused octahydronaphthalene nucleus by compari-
1
son if its H NMR data with that previously obtained for 7.34
The resonance of H(7) (δ 4.10, J ) 2.8 Hz) observed for 45 is
characteristic of cis-fused octahydronaphthalenes bearing an
axial C(7) alkoxy substituent.15,34 Cycloadduct 45, like 44,
displayed characteristic 1H signals that define the top half
cyclohexenyl unit as deriving from an exo transition state (e.g.,
H(18), δ 3.14 (d, J ) 8.8 Hz); H(22ax), δ 2.12 (dd, J ) 14.0,
7.2 Hz); and C(21)-Me, δ 1.13 (d, J ) 7.2 Hz)).
Further HPLC separation of mixed fractions provided very
small amounts of an inseparable 1:1 mixture of two minor
1
double cycloadducts. The H NMR spectrum of this mixture
contained a signal at δ 3.14 (d, J ) 8.8 Hz) corresponding to
H(18), characteristic of an exo top half and also a multiplet at
δ 3.03 (apparent triplet) that is characteristic of H(18) of an
endo-Diels-Alder derived top half.37,53,56 The 1H NMR
spectrum of this mixture also exhibited a multiplet at δ 3.23
(dt, J ) 4.0, 10.0 Hz) indicative of an equatorial MOM ether
on a trans-fused octahydronaphthalene nucleus (as in 44), as
well as multiplets at δ 1.80-1.90 that have the same shape and
appearance as those in the spectrum of the axial MOM trans-
fused adduct 8.15,34 Accordingly, structures of the two cycload-
ducts in this minor 1:1 mixture have tentatiVely been assigned
as 46 with an endo top half and a trans-fused bottom half with
an equatorial MOM ether, and 47 with an exo top half and a
trans-fused bottom half with an axial MOM ether. These
assignments also correlate with the predicted product distribution
based on earlier studies on the Diels-Alder reactions of 5 and
30.
The stereostructures of 44 and 24 were assigned by correlation
1
of H NMR data with NMR data previously obtained for 634
and 31.37 In particular, both 44 and 24 displayed J3,8 ) 8.8-
9.4 Hz and J7,8 ) 10 Hz, characteristic of trans-fused octahy-
dronaphthalenes with equatorial C(7) alkoxy substituents (δ 3.13
in both structures), while 44 displayed resonances for H(18) at
δ 3.13 (d, J ) 8.4 Hz), H(22ax) at δ 2.11 (dd, J ) 14.0, 7.2
Hz), and C(21)-Me at δ 1.13 (d, J ) 7.2 Hz). The latter data
are characteristic of a top half Diels-Alder adduct deriving from
an exo transition state.37,53,56
This yield of 44 is remarkably close to the maximum yield
(67%) anticipated on the assumption that the bimolecular Diels-
Alder reaction of (R)-12 and the C(16)-C(21) triene unit of 19
should proceed with 93:7 selectivity, as modeled by the reaction
One additional (also inseparable) 1:1 mixture of two minor
products was obtained. One of the products in this mixture
was tentatively assigned (based on 1H NMR data) as 48
containing a cis-fused bottom half and an intact C(16)-C(21)

Total Synthesis of (-)-Chlorothricolide
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7417
triene unit.82 The second component of this mixture contains
HMPA were added to provide the fully protected tetronic ester
51 in 94% yield.13,53 The two allyl protecting groups were then
removed in a single step by treatment of 51 with catalytic
Ph(PPh3)4 and dimedone in THF.84 This provided the seco acid
52 in 94% yield (88% from 50).
an exo top half, owing to the characteristic 1H signal at δ 3.14
(d, J ) 8.8 Hz) for H(18), but with an unknown stereochemistry
of the bottom half. Attempts to separate these mixtures for
further stereochemical analysis were unsuccessful.
Considerable effort was devoted to the optimization of the
tandem Diels-Alder reaction and the recycle of the intra-
molecular adduct (E,E,E)-24. The most consistent results were
obtained when hexaenoate 19 was heated with 2 equiv of (R)-
12 at 120 °C in toluene (1 M). The yield and ratio of products
were fairly constant when the reaction was performed at
concentrations between 1.0 and 2.0 M. However, at higher
reaction temperatures or concentrations, or when greater than
2 equiv of (R)-12 was used, polymerization of the reaction
mixture began to be a significant problem, leading to substan-
tially reduced yields of product. Longer reaction times (>24
h) did not lead to improved yields of 44, but did lead to reduced
isolated yields of recovered (E,E,E)-24. These observations
suggest that the stabilities of (R)-12 and (E,E,E)-24 are the
factors that limit the overall efficiency of the tandem Diels-
Alder reaction.
The recycle reaction of intramolecular cycloadduct (E,E,E)-
24 and dioxolanone (R)-12 was performed most efficiently at
125 °C in trichloroethylene. This reaction required somewhat
higher concentrations (1.5 M) and longer reaction times (28-
48 h) to achieve maximum efficiency. Unfortunately, the
recycle reaction was also constrained by a polymerization
problem that resulted with increased reaction temperature (>130
°C) or when greater quantities (>2.0 equivalents) of dienophile
were used. Nevertheless, under optimal conditions, the Diels-
Alder reaction of (E,E,E)-24 with 2 equiv of (R)-12 in
trichloroethylene provided 44 in up to 58% yield, along with
recovered (E,E,E)-24 (20-45%) that could be recycled further.
As already noted, the yield of the double Diels-Alder adduct
44 was 55-59% from 19 after one such recycle.
The simplicity of the four step conversion of 44 to 52 belies
the considerable difficulty encountered in developing it. Prior
to the identification of 51 as a suitable precursor to the seco
acid, compounds 53-55 were also prepared. However, attempts
Elaboration of 44 to (-)-Chlorothricolide. Intermediate
44 was elaborated to the seco acid 52 as summarized below.
First, treatment of 44 with K2CO3 in MeOH provided the
corresponding R-hydroxy methyl ester, which was then esterified
with allyloxy acetic acid 4983 in the presence of DCC and
DMAP,81 thereby providing triester 50 in 91% overall yield.
Dieckmann closure of the spiro tetronate was accomplished by
treatment of 50 with LiN(TMS)2 in THF at -78 °C. The enolate
solution was allowed to warm to 0 °C and then MOM-Cl and
to deprotect the tetronate 2-(trimethylsilyl)ethoxymethyl (SEM)
ether by treatment of 53 with TBAF in THF resulted in total
decomposition of the starting material. Similarly, attempts to
remove the p-methoxybenzyl (PMB) ether protecting group from
the tetronate unit of 54 ((DDQ, wet CH2Cl2;85 or CAN, wet
CH3CN86), were unsuccessful, owing to the instability of the
product 56 toward these mild oxidants.87 Ultimately, we
identified the allyl ether protecting group88 as the most ap-
propriate for the enolic C(25) tetronate hydroxyl group, in
(82) Cycloadduct 48 was also obtained as the second most predominant
product of the IMDA reaction of 19 performed (125 °C, 24 h, toluene,
BHT inhibitor) in the absence of (R)-12. The major product of this reaction,
45, was obtained in 40-47% yield following HPLC purification, while a
mixture of 48 and a third IMDA product (presumably corresponding to
47) was obtained in 13-17% yield.
(84) Kunz, H.; Waldmann, H. Angew. Chem., Int. Ed. Engl. 1984, 23,
71.
(85) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O.
Tetrahedron 1986, 42, 3021.
(86) Johansson, R.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1 1984,
2371.
(87) A sample of 56, prepared by deprotection of the allyl ether of 55,
decomposed when exposed to DDQ in wet CH2Cl2.
(83) Manhas, M.; Amin, S.; Chawla, H. P. S.; Bose, A. K. J. Heterocycl.
Chem. 1978, 15, 601.

7418 J. Am. Chem. Soc., Vol. 120, No. 30, 1998
Roush and Sciotti
anticipation that this unit should be more acidic than most
alcohols and have reactivity comparable to a phenol. The
facility of the reaction of the C(25) allyl ether with Pd(0) and
dimedone attests to the soundness of these arguments. Indeed,
treatment of 55 with catalytic Pd(PPh3)4 and stoichiometric
dimedone provided 56 in 83% yield. However, attempts to
unmask the C(1) carboxylic acid unit of 56 also were thwarted.
Treatment of 56 with a number of fluoride sources (e.g., KF in
DMSO;89 CsF in DMF;90 or TBAF in THF, DMF, or CH3CN91)
resulted in decomposition or no reaction. Fortunately, the allyl
unit that solved the tetronate C(25) OH protecting group problem
also provided a workable solution to the C(1) acid protecting
group issue (vide supra). An additional benefit to use of the
allyl protecting group scheme for 51 is that both allyl groups
could be removed simultaneously, thereby further simplifying
the synthetic sequence.
CH3CN at 50 °C removed the primary TBDPS ether.99 Oxida-
tion of the allylic alcohol with MnO2 and then oxidation of the
aldehyde to the carboxylic acid by using NaClO2 in aqueous
t-BuOH in the presence of isobutylene provided 58 in 75% yield
for the three steps.100,101 Finally, simultaneous removal of the
two MOM ethers and cleavage of the vinylsilane unit by
treatment of 58 with EtSH and BF3‚Et2O102 provided synthetic
(-)-chlorothricolide, (-)-(1) ([R]25 ) -23° (c ) 0.2, CH2-
D
Cl2)) in 91% yield. Because an authentic reference sample of
(-)-chlorothricolide was not available, synthetic (-)-1 was
converted to 24-O-methylchlorothricolide methyl ester (59), a
compound obtained in the original structural work on chloro-
thricin3 and an intermediate in Yoshii’s synthesis of racemic
24-O-methylchlorothricolide.31 Thus, synthetic (-)-chlorothri-
With the seco acid 52 in hand, our attention turned to the
macrolactonization reaction.92 This transformation was best
accomplished by treating 52 with bis(2-oxo-3-oxazolidinyl)-
phosphonic chloride (BOP-Cl) (1.9 equiv) and Et3N (3.9 equiv)
at 100 °C for 1 h in toluene (0.01 M), which provided
macrocycle 57 in 50% yield along with 31% of recovered seco
acid 52.93 Attempts to improve the efficiency of this reaction
by using additional equivalents of BOP-Cl (up to 8 equiv),
lower reaction temperatures, longer reaction times, or by
performing the macrolactonization in the presence of activated
4 Å molecular sieves were unsuccessful. Macrocycle 57 was
obtained in low yield (10-20%) using the Yamaguchi lacton-
ization (trichlorobenzoyl chloride, DMAP).94,95 Use of DCC
and DMAP gave the N-acyl urea rather than 57, while use of
DCC, DMAP, and DMAP‚HCl gave a very low yield of the
macrocycle.96 Similarly, only traces of 57 were obtained from
experiments performed using CDI97 and DBU or trifluoroacetic
anhydride and Et3N98 as the dehydrating agents.
The synthesis was completed by a simple four-step sequence,
as follows. First, treatment of lactone 57 with Et3N‚HF in
colide was treated with excess diazomethane in diethyl ether to
give 24-O-methylchlorothricolide methyl ester 59 [([R]25
D
(88) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic
Synthesis, 2nd ed.; Wiley-Interscience: New York, 1991.
(89) Roush, W. R.; Blizzard, T. A. J. Org. Chem. 1984, 49, 1772.
(90) Suzuki, K.; Matsumoto, T.; Tomooka, K.; Matsumoto, K.; Tsuchi-
hashi, G.-i. Chem. Lett. 1987, 113.
-29.3° (c ) 0.95, CHCl3), mp 228.5-229 °C; lit.3 [R]20D -30°
(c ) 1, CHCl3); lit.3 mp 230 °C)] in 80% yield. Synthetic (-)-
59 was identical in all respects (except optical rotation and
melting point) with an authentic sample of racemic 59.31
Summary. This work constitutes the first total synthesis of
(-)-chlorothricolide (1), the aglycon of the antibiotic chloro-
thricin. The synthesis proceeds in about 2% overall yield and
20 steps via the longest linear sequence originating from the
known acetylenic ketone 33. The synthesis features the tandem
(91) Sieber, P. HelV. Chim. Acta 1977, 60, 2711.
(92) Meng, Q.; Hesse, M. Topics Curr. Chem. 1991, 161, 107.
(93) Corey, E. J.; Hua, D. H.; Pan, B.-C.; Seitz, S. P. J. Am. Chem. Soc.
1982, 104, 6818.
(94) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull.
Chem. Soc. Jpn. 1979, 52, 1989.
(95) Hikota, M.; Tone, H.; Horita, K.; Yonemitsu, O. J. Org. Chem.
1990, 55, 7.
(96) Boden, E. P.; Keck, G. E. J. Org. Chem. 1985, 50, 2394.
(97) Colvin, E. W.; Purcell, T. A.; Raphael, R. A. J. Chem. Soc., Perkin
Trans. 1 1976, 1718.
(98) Taub, D.; Girotra, N. N.; Hoffsomer, R. D.; Kuo, C. H.; Slates, H.
L.; Weber, S.; Wendler, N. L. Tetrahedron 1968, 24, 2443.
(99) Hu¨nig, S.; Wehner, G. Synthesis 1975, 180.
(100) Lindgren, B. O.; Nilsson, T. Acta Chem. Scand. 1973, 27, 888.
(101) Kraus, G. A.; Taschner, M. J. J. Org. Chem. 1980, 45, 1175.
(102) Fuji, K.; Ichikawa, K.; Node, M.; Fujita, E. J. Org. Chem. 1979,
44, 1661.

Total Synthesis of (-)-Chlorothricolide
J. Am. Chem. Soc., Vol. 120, No. 30, 1998 7419
inter-intramolecular Diels-Alder reaction of hexaenoate 19 and
the chiral dienophile (R)-12 that establishes seven of the eight
stereocenters of (-)-chlorothricolide in a single operation, with
an overall stereoselectivity of ca. 67%. It is important to stress
that the tandem Diels-Alder reaction strategy is feasible only
by virtue of the exceptionally high diastereofacial and exo
selectivity of the chiral dienophile 12, which enables the relative
and absolute stereochemistry of the three stereocenters of the
top half fragment to be established independent of the five
stereocenters in the bottom half perhydronaphthalene unit. The
C(9)-trimethylsilyl substituent of 19 also plays a key strategic
role by serving as the stereochemical control element for the
IMDA reaction leading to the bottom half octahydronaphthalene
unit.103
Acknowledgment. This research was supported by a grant
from the National Institute of General Medical Sciences (GM
26782). We also thank Professor Yoshii for providing a
reference sample of racemic 59 for comparative purposes.
Supporting Information Available: Complete experimental
1
procedures for the total synthesis of (-)-chlorothricolide, H
NMR spectra of (-)-1, 19, 21, 22, (E,E,E)-24, 41, 42, 44, 45,
52, 58, and 59, and 13C NMR spectra of 59 (31 pages, print/
PDF). See any current masthead page for ordering information
and Web access instructions.
(103) Complete experimental details are provided in the Supporting
Information.
JA980611F