Total synthesis of (-)-bafilomycin A1

Article abstract of DOI:10.1021/ja017885e

A highly stereoselective total synthesis of (-)-bafilomycin A₁, the naturally occurring enantiomer of this potent vacuolar ATPase inhibitor, is described. The synthesis features the highly stereoselective aldol reaction of methyl ketone 8b and

Full text of DOI:10.1021/ja017885e

Published on Web 05/23/2002
Total Synthesis of (-)-Bafilomycin A1
Karl A. Scheidt,1 Thomas D. Bannister,2 Akihiro Tasaka,2 Michael D. Wendt,2
Brad M. Savall,1 Glenn J. Fegley,2 and William R. Roush*,1
Contribution from the Department of Chemistry, UniVersity of Michigan,
Ann Arbor, Michigan 48109, and Department of Chemistry, Indiana UniVersity,
Bloomington, Indiana 47405
Received December 27, 2001
Abstract: A highly stereoselective total synthesis of (-)-bafilomycin A1, the naturally occurring enantiomer
of this potent vacuolar ATPase inhibitor, is described. The synthesis features the highly stereoselective
aldol reaction of methyl ketone 8b and aldehyde 60c and a Suzuki cross-coupling reaction of the highly
functionalized advanced intermediates 12 and 39. Vinyl iodide 12 was synthesized by a 14-step sequence
starting from the readily available â-alkoxy aldehyde 14, while the vinylboronic acid component 39 was
synthesized by a nine-step sequence from â-hydroxy-R-methyl butyrate 44 via a sequence involving the
R-methoxypropargylation of chiral aldehyde 49 with the R-methoxypropargylstannane reagent 54. Syntheses
of fragments 12 and 39 also feature diastereoselective double asymmetric crotylboration reactions to set
several of the critical stereocenters. The Suzuki cross-coupling of 12 and 39 provided seco ester 40, which
following conversion to the seco acid underwent smooth macrolactonization to give 41. The success of the
macrocyclization required that C(7)-OH be unprotected. The Mukaiyama aldol reaction between aldehyde
60c and the TMS enol ether generated from 8b provided aldol 65 with high diastereoselectivity. Finally, all
silicon protecting groups were removed by treatment of the penultimate intermediate 65 with TAS-F (tris-
(dimethylamino)sulfonium difluorotrimethylsilicate), thereby completing the total synthesis of (-)-bafilomycin
A1.
Introduction
that is necessary for biological activity. The structurally distinc-
tive C(2)-C(5) dienyl methyl ether system is also found in other
members of this family.
Bafilomycin A1 (1)3 is a member of the plecomacrolide family
(formerly known as the hygolide family)4 of macrolide antibiot-
ics that includes the hygrolidins,5,6 the concanamycins (e.g.,
concanomycin A, 2),7 and formamicin, 3.8,9 Bafilomycin A1 is
a potent vacuolar H+-ATPase inhibitor that displays broad
antibacterial and antifungal activity.10 The stereochemistry of
bafilomycin A1, initially assigned by Corey on the basis of a
molecular modeling analysis of published NMR data,4 was
subsequently verified by X-ray crystallography.11,12 Bafilomycin
A1 contains an acid- and base-sensitive six-membered hemiketal
that participates in a hydrogen-bond network with the C(17)
hydroxyl group and the carbonyl of the 16-membered lactone
In view of the interesting chemical structures and potent
biological properties, considerable effort has been devoted to
the development of efficient syntheses of members of this
family. Total syntheses of bafilomycin A1 have been recorded
by Evans13,14 and Toshima,15-17 and very recently total syntheses
of bafilomycin A1 and V1 have been accomplished by Hanessian
and Marshall, respectively.18,19 A total synthesis of hygrolidin
has been accomplished by Yonemitsu,20,21 and total syntheses
of concanamycin F (the aglycone of concanamycin A) have been
recorded by both the Toshima and Paterson groups.22-24 Several
25-28
studies on the synthesis of bafilomycin A1
and of bafilo-
(1) University of Michigan.
(2) Indiana University.
(3) Werner, G.; Hagenmaier, H.; Drautz, H.; Baumgartner, A.; Za¨hner, H. J.
Antibiot. 1984, 37, 110.
(13) Evans, D. A.; Calter, M. A. Tetrahedron Lett. 1993, 34, 6871.
(14) Calter, M. A. Ph.D. Thesis, Harvard University, 1993.
(15) Toshima, K.; Jyojima, T.; Yamaguchi, H.; Murase, H.; Yoshida, T.;
Matsumura, S.; Nakata, M. Tetrahedron Lett. 1996, 37, 1069.
(16) Toshima, K.; Yamaguchi, H.; Jyojima, T.; Noguchi, Y.; Nakata, M.;
Matsumura, S. Tetrahedron Lett. 1996, 37, 1073.
(4) Corey, E. J.; Ponder, J. W. Tetrahedron Lett. 1984, 25, 4325.
(5) Seto, H.; Tajima, I.; Akao, H.; Furihata, K.; Otake, N. J. Antibiotics 1984,
37, 610.
(6) Seto, H.; Akao, H.; Furihata, K.; Otake, N. Tetrahedron Lett. 1982, 23,
2667.
(17) Toshima, K.; Jyojima, T.; Yamaguchi, H.; Noguchi, Y.; Yoshida, T.;
Murase, H.; Nakata, M.; Matsumura, S. J. Org. Chem. 1997, 62, 3271.
(18) Hanessian, S.; Ma, J.; Wang, W. J. Am. Chem. Soc. 2001, 123, 10200.
(19) Marshall, J. A.; Adams, N. D. J. Org. Chem. 2002, 67, 733.
(20) Makino, K.; Kimura, K.; Nakajima, N.; Hashimoto, S.; Yonemitsu, O.
Tetrahedron Lett. 1996, 37, 9073.
(7) Kinashi, H.; Someno, K.; Sakaguchi, K.; Higashijima, T.; Miyazawa, T.
Tetrahedron Lett. 1981, 22, 3861.
(8) Igarashi, M.; Kinoshita, N.; Ikeda, T.; Nakagawa, E.; Hamada, M.; Takeuchi,
T. J. Antibiot. 1997, 50, 926.
(9) Igarashi, M.; Nakamura, H.; Naganawa, H.; Takeuchi, T. J. Antibiot. 1997,
50, 932.
(21) Makino, K.; Nakajima, N.; Hashimoto, S.; Yonemitsu, O. Tetrahedron Lett.
1996, 37, 9077.
(10) Bowman, E. J.; Siebers, A.; Altendorf, K. Proc. Natl. Acad. Sci. U.S.A.
1988, 85, 7972.
(22) Jyojima, R.; Miyamoto, N.; Katohno, M.; Nakata, M.; Matsumura, S.;
Toshima, K. Tetrahedron Lett. 1998, 39, 6007.
(11) Baker, G. H.; Brown, P. J.; Dorgan, R. J. J.; Everett, J. R.; Ley, S. V.;
Slawin, A. M. Z.; Williams, D. J. Tetrahedron Lett. 1987, 28, 5565.
(12) Baker, G. H.; Brown, P. J.; Dorgan, R. J. J.; Everett, J. R. J. Chem. Soc.,
Perkin Trans. 2 1989, 1073, and references therein.
(23) Toshima, K.; Jyojima, T.; Miyamoto, N.; Katohno, M.; Nakata, M.;
Matsumura, S. J. Org. Chem. 2001, 66, 1708.
(24) Paterson, I.; Doughty, V. A.; McLeod, M. D.; Trieselmann, T. Angew.
Chem., Int. Ed. 2000, 39, 1308.
9
10.1021/ja017885e CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 6981-6990
6981

A R T I C L E S
Scheidt et al.
Scheme 1. Key Subunits of Bafilomycin A1
mycin analogues29-31 have also appeared. We report the details
of our total synthesis of bafilomycin A1, a preliminary account
of which was published in 1999.32
segment could be synthesized with 8:1 selectivity (for 7a), via
the aldol reaction between 5 and the lithium enolate of 6a, or
with g 16:1 selectivity (for 7b or 7c), by the coupling of 5 and
the chlorotitanium enolates of 6b or 6c. We also found that the
stereoselectivity of the lithium enolate aldol couplings was
strikingly dependent on the C(23) protecting group of the
aldehyde 5, as well as on the nature of the functionality at C(15)
of the methyl ketone fragment. Most troublesome, however, was
our inability to devise a suitable means of protecting the C(19)-
carbonyl group of intermediates such as 7, which would be
required if 7 were to be used as an intermediate for further
elaboration to the natural product. Accordingly, we decided to
postpone the aldol reaction between 5 and the C(20) methyl
ketone until after fragments 4 and 6 were coupled via the
C(10)-C(14) diene. We recognized that it would be preferable
to postpone the aldol step until after the bafilomycin macrocycle
was assembled, since this approach would minimize protecting
group manipulations of the C(15)-hydroxyl group (Scheme 1).
On the basis of this analysis, we targeted the fully elaborated
macrocyclic methyl ketone 8 as a key synthetic intermediate.
We envisaged that 8 could be assembled via macrocyclization
of the seco ester precursor 934 and that 9 in turn could be
generated by coupling of suitably functionalized fragments
deriving from 4 and precursors to 6. Two distinct olefination
sequences were envisaged. One involved the Horner35 coupling
of 10a and aldehyde 11 or, equivalently, the modified Julia
coupling of 11 and the allylic sulfone 10b.36 The second strategy
would utilize the Suzuki cross coupling reaction of vinyl iodide
12 and vinylboronic acid 13 (Scheme 2).37
Chemistry
Evolution of the Synthetic Strategy. From the outset, our
strategy focused on the synthesis of bafilomycin A1 from three
key subunits: 4, corresponding to the C(5)-C(11) segment of
the natural product; aldehyde 5, the C(21)-C(25) fragment; and
methyl ketone 7, the C(13)-C(20) unit. In studies described in
detail elsewhere,27,28,33 we established that the C(13)-C(25)
Both of these olefination strategies were explored during the
course of this synthesis. However, despite extensive efforts to
implement the Horner and the Julia olefination sequences, these
approaches were ultimately abandoned, since we were unable
to devise a high-yielding and highly stereoselective synthesis
of the targeted (E,E)-diene system by using these reactions; a
summary of these efforts is provided in the Supporting Informa-
tion. Consequently, we focus here on the Suzuki cross-coupling
strategy for completion of the bafilomycin A1 total synthesis.
(25) Paterson, I.; Bower, S.; McLeod, M. D. Tetrahedron Lett. 1995, 36, 175.
(26) Marshall, J. A.; Adams, N. D. Org. Lett. 2000, 2, 2897.
(27) Roush, W. R.; Bannister, T. D. Tetrahedron Lett. 1992, 33, 3587.
(28) Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993, 34,
8387.
(29) Hanessian, S.; Tehim, A.; Meng, Q.; Grandberg, K. Tetrahedron Lett. 1996,
37, 9001.
(33) Roush, W. R.; Bannister, T. D.; Wendt, M. D.; Jablonowski, J. A.; Scheidt,
K. A. J. Org. Chem. 2002, in press.
(30) Gatti, P. A.; Gagliardi, S.; Cerri, A.; Visconti, M.; Farina, C. J. Org. Chem.
1996, 61, 7185.
(34) Meng, Q.; Hesse, M. Top. Curr. Chem. 1991, 161, 107.
(35) Maryanoff, B. E.; Reitz, A. B. Chem. ReV. 1989, 89, 863.
(36) Baudin, J. B.; Hareau, G.; Julia, S. A.; Lorne, R.; Ruel, O. Bull. Soc. Chim.
Fr. 1993, 130, 856.
(31) Gagliardi, S.; Gatti, P. A.; Belfiore, P.; Zocchetti, A.; Clarke, G. D.; Farina,
C. J. Med. Chem. 1998, 41, 1883.
(32) Scheidt, K. A.; Tasaka, A.; Bannister, T. D.; Wendt, M. D.; Roush, W. R.
Angew. Chem., Int. Ed. Engl. 1999, 38, 1652.
(37) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457.
9
6982 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Total Synthesis of (−)-Bafilomycin A
A R T I C L E S
1
Scheme 2. Key Synthetic Intermediates
Scheme 3. Synthesis of Vinyl Iodide 12
Transition metal-mediated cross-coupling reactions have been
used in many natural product syntheses.37,38 The Stille reaction,
in particular, has proven to be extremely useful for the synthesis
of structurally complex intermediates and targets. Indeed,
Evans,13,14 Toshima,17 and recently also Hanessian18 employed
Stille cross-coupling reactions to generate the (E,E)-dienes in
their syntheses of bafilomycin A1. In contrast, there are far fewer
applications of the Suzuki reaction for the late-stage union of
complex intermediatessthe most notable of the limited examples
being those in Kishi’s palytoxin synthesis39 and Evans’ ruta-
mycin B synthesis.40 The cross coupling reactions of 12 and 13
reported herein thus expands the number of successful applica-
tions of the Suzuki cross-coupling reaction for the late-stage
union of highly functionalized intermediates in the total synthesis
of complex natural products.
Synthesis of Vinyl Iodide 12. Vinyl iodide 12 was synthe-
sized starting from the known aldehyde 1441 by a route that
(39) Armstrong, R. W.; Beau, J.-M.; Cheon, S. H.; Christ, W. J.; Fujioka, H.;
Ham, W.-H.; Hawkins, L. D.; Jin, H.; Kang, S. H.; Kishi, Y.; Martinelli,
M. J.; W. W. McWhorter, J.; Mizuno, M.; Nakata, M.; Stutz, A. E.;
Talamas, F. X.; Taniguchi, M.; Tino, J. A.; Ueda, K.; Uenishi, J.-i.; White,
J. B.; Yonaga, M. J. Am. Chem. Soc. 1989, 111, 7525.
(38) Farina, V.; Krishnamurthy, V.; Scott, W. J. Org. React. 1997, 50, 1.
(40) Evans, D. A.; Ng, H. P.; Rieger, D. L. J. Am. Chem. Soc. 1993, 115, 11446.
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 6983

A R T I C L E S
Scheidt et al.
features the synthesis of the challenging anti-anti stereotriad42,43
by using our diastereoselective aldehyde crotylboration technol-
ogy44 in the mismatched manifold (Scheme 3).45 Thus, the
double asymmetric reaction of 14 and (R,R)-diisopropyl tartrate
(E)-crotylboronate (15) provided an 85:15 mixture of 16 and
its 3,4-anti-4,5-syn diastereomer (77% isolated yield of 16). The
free hydroxyl group of 16 was protected as a tert-butyldimeth-
ylsilyl (TBS) ether, and then the vinyl group was converted to
the primary alcohol in 17 via Rh(I)-catalyzed hydroboration with
catecholborane.46 Oxidation of 17 using the standard Moffatt-
Swern conditions47 provided the corresponding aldehyde that
was then transformed smoothly to alkyne 18 by using the
Corey-Fuchs protocol.48 The p-methoxybenzyl (PMB) ether
was removed (96%) by using the standard DDQ protocol,49 and
then the alkynol 19 was converted to vinyl iodide 20 using
Negishi’s carbozirconation methodology (65%).50 It was neces-
sary to use 1,2-dichloroethane as the reaction solvent, as the
reaction was considerably slower and gave significantly dimin-
ished yields when performed in CH2Cl2 at reflux (40 °C).
Interestingly, the Negishi reaction also proved to be very
sensitive to the presence of a free hydroxy group in the substrate.
The reaction was entirely unsuccessful using 18 or 25 in which
both hydroxyl groups were fully protected. Diol 26 proved to
be the best substrate and gave the vinyl iodide 27 in 91% yield.17
However, intermediate 27 was not pursued, owing to the
additional steps that would be required to differentiate the two
hydroxyl groups for use of this compound in the bafilomycin
synthesis.
of this unit in his concanamycin A synthesis.51 Selectivity for
24 was only ca. 2:1 when the corresponding dimethylphospho-
nate reagent was employed. Because initial macrolactonization
experiments indicated that the C(7) TBS ether prevented
cyclization (vide infra), the offending TBS group was removed
by treatment of 24 with TBAF to generate vinyl iodide 12
(82%).
Synthesis of Vinylboronic Acid 13. Vinylboronic acid 13
was synthesized starting from homoallylic TES ether 28, which
we had initially prepared during our studies on the fragment
assembly aldol reaction of 5 and 6 en route to the C(13)-C(25)
segment of the natural product.33 Oxidative cleavage of the vinyl
group to give aldehyde 29 was accomplished by olefin dihy-
droxylation followed by cleavage of the diol with Pb(OAc)4.
Oxidative cleavage of the diol with NaIO4 resulted in epimer-
ization at C(14) of sensitive R-methoxy aldehyde 29; epimer-
ization at C(14) was also observed when 29 was prepared by
ozonolysis of 28. The fully protected aldehyde 29 was smoothly
converted to the corresponding alkyne 30 by using the Gilbert-
Seyferth R-diazomethylphosphonate reagent.52,53 The synthesis
of vinylboronic acid 13 was then completed by hydroboration
of 30 with catecholborane.
Vinyl iodide 20 was elaborated to enoate 21 in 90% yield
via Moffatt-Swern oxidation and stabilized Wittig olefination
using Ph3PdC(Me)CO2Me in toluene at 60 °C. DIBAL reduc-
tion of the ester gave the allylic alcohol, which was oxidized
to the enal 22 by using MnO2 (97% for the two steps). Finally,
the dienoate unit was introduced with greater than 95:5 (Z,E:
E,E)-selectivity (85% yield) by using a Horner-Wadsworth-
Emmons reaction with (i-PrO)2P(O)CH(OMe)CO2Me (23) in
the presence of KN(SiMe3)2 (KHDMS) and 18-crown-6 in THF,
using the conditions developed by Paterson for the installation
Fragment Assembly and MacrocyclizationsInitial Stud-
ies. The Suzuki coupling of vinyl iodide 12 and vinylboronic
acid 13 was performed by using TlOH as the base and provided
the (E,E)-diene 31 in 80% yield.54 Deprotection of the C(15)-
TES ether by using trifluoroacetic acid (TFA) in THF/water
provide the seco ester in 85% yield. Saponification of the methyl
ester with 1 N KOH in dioxane then provided the unpurified
seco acid that was treated with 20 equiv of 2,4,6-trichloroben-
zoyl chloride and 40 equiv of i-Pr2NEt in THF.55 Large excesses
of both the acid chloride and base were necessary to prevent
symmetrical anhydride formation, which we have isolated and
found to be relatively unreactive under the macrocyclization
conditions. This phenomenon has also been observed in the
hygrolide synthesis by Yonemitsu.21 The THF used in the
formation of the mixed anhydride was removed and replaced
(41) Smith, A. B., III; Qiu, Y.; Jones, D. R.; Kobayashi, K. J. Am. Chem. Soc.
1995, 117, 12011.
(42) Hoffmann, R. W.; Dahmann, G.; Andersen, M. W. Synthesis 1994, 629.
(43) Hoffmann, R. W.; Stahl, M.; Schopfer, U.; Frenking, G. Chem.-Eur. J.
1998, 4, 559.
(44) Roush, W. R.; Palkowitz, A. D.; Ando, K. J. Am. Chem. Soc. 1990, 112,
6348.
(45) Masamune, S.; Choy, W.; Petersen, J. S.; Sita, L. R. Angew. Chem., Int.
Ed. Engl. 1985, 24, 1.
(46) Evans, D. A.; Fu, G. C.; Hoveyda, A. H. J. Am. Chem. Soc. 1992, 114,
6671.
(51) Paterson, I.; McLeod, M. D. Tetrahedron Lett. 1997, 38, 4183.
(52) Gilbert, J. C.; Weerasooriya, U. J. Org. Chem. 1982, 47, 1837.
(53) Brown, D. G.; Velthuisen, E. J.; Commerford, J. R.; Brisbois, R. G.; Hoye,
T. R. J. Org. Chem. 1996, 61, 2540.
(54) Uenishi, J.-i.; Beau, J.-M.; Armstrong, R. W.; Kishi, Y. J. Am. Chem. Soc.
1987, 109, 4756.
(55) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Chem. Bull.
Jpn. 1979, 52, 1989.
(47) Tidwell, T. T. Org. React. 1990, 39, 297.
(48) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769.
(49) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O.
Tetrahedron 1986, 42, 3021.
(50) Negishi, E.-i.; Van Horn, D. E.; Yoshida, T. J. Am. Chem. Soc. 1985, 107,
6639.
9
6984 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Total Synthesis of (−)-Bafilomycin A
A R T I C L E S
1
with benzene, DMAP was added, and the resulting solution was
heated to reflux for 15 h, thereby providing macrolactone 32 in
45% yield from the seco ester (36% yield from 31).
Another issue concerning protecting groups was immediately
encountered during attempts to remove the dimethoxybenzyl
ether (DMPM) protecting group from macrolactone 32: all
conditions explored to unmask the C(19) hydroxy group of 32
promoted substantial decomposition (DDQ, wet CH2Cl2; Ce-
(NH4)4(NO3)6; MgBr2‚OEt2, Me2S, etc.). We presume that allylic
oxidation of the dienylic methyl ether unit is a facile process
with strong oxidants such as DDQ or CAN58 and that nucleo-
philic conditions are harmful to the macrocycle. It is interesting
to note that attempted DDQ deprotection of seco ester 31 also
resulted in decomposition, while deprotection of the DMPM
ether of allylic methyl ether 30 proceeds in 94% yield.
Critical to the success of this macrolactonization is the fact
that the C(7)-hydroxyl is unprotected in the lactonization
substrate. Earlier attempts to effect macrolactonization of seco
acids 9 or 33 with C(7)-TBS ethers were largely unsuccessful.56
In one case (cyclization of 9, R ) DMPM, using the Yamaguchi
mixed anhydride protocol),55 the 16-membered lactone was
obtained in ca. 10% yield. However, this result could not be
reproduced, and no lactone was obtained under any of the other
conditions examined.57
Revisions of the Protection Group Scheme. Our experience
with attempted deprotections of macrocycle 36 and its seco ester
precursor 31 necessitated that the offending DMPM group be
removed before the macrocylization. Because free hydroxyl
groups are fully compatible with the Suzuki cross coupling
protocol, we anticipated that it might be possible to proceed
with C(19)-OH unprotected. Accordingly, the DMPM group of
alkynyl methyl ether 30 was removed using DDQ and the C(15)-
OTES ether was hydroylzed under acidic conditions, giving diol
34 in 83% yield. This intermediate was hydroborated using
catecholborane and a catalytic amount of 9-BBN, which
provided vinylboronic acid 35 after aqueous workup and column
chromatographic purification.59
The Pd(0)-mediated cross-coupling of vinylboronic acid 35
and vinyl iodide 12 proceeded smoothly, giving the seco ester
36 in 60% yield (unoptimized). We hoped that the macrocy-
clization of the seco acid derived from 36 would close
preferentially on C(15)-OH to give the 16-membered macrolide
38, rather than on C(19), which would provide the 20-membered
macrocycle 37. In the event, however, saponification of 36 and
cyclization of the resulting seco acid using the modified
Yamaguchi conditions provided the undesired 20-membered
lactone 37 in 43% yield, along with only 7% of the desired
16-membered macrocycle 38. We also were unable to develop
conditions to effect isomerization of 37 to 38, although ring
contractions of this type have been demonstrated previously in
both the iso-bafilomycin and scytophycin series.60,61
Synthesis of Macrocylic Methyl Ketone 8b. It was clear
from the results with 31 and 36 that the macrocyclization
substrate must have an easily removable protecting group at
C(19). Accordingly, we elected to protect the C(19)-OH as a
triethylsilyl (TES) ether. Thus, the less hindered C(19)-hydroxyl
of diol 34 was selectively protected by treatment with TES-Cl
and pyridine (92% yield). Hydroboration of the alkyne using
Inspection of the X-ray crystal structure11,12 of bafilomycin
A1 reveals that the C(7)-hydroxyl is axial with respect to the
plane of the relatively flat 16-membered macrocycle. Moreover,
inspection of molecular models indicated that the C(7)-TBS
ether protecting group experiences destabilizing interactions with
the pseudoequatorial C(6) methyl or the C(9) methylene groups,
depending on the conformation of the C(7)-OSiMe2t-Bu unit.
Accordingly, we presume that the C(7)-OTBS unit influences
the conformational distribution of seco acid 33, such that the
precyclization conformer required for macrocyclization is less
significantly populated than is the case with the seco acid
precursor to 32, which lacks the offending C(7)-OTBS group.
This analysis is consistent with Evans’ observation that the
macrocylization of his bafilomycin seco acid containing a C(7)-
OTBS protecting group took 7 times longer (72 h) than the
cyclization of the seco acid where the C(7)-OH is unprotected
(10 h).14
(58) Examples of DDQ oxidation of allylic PMB ethers that give enone
products: (a) Kanda, Y.; Fukuyama, T. J. Am. Chem. Soc. 1993, 115, 6451.
(b) Trost, B. M.; Chung, J. Y. J. Am. Chem. Soc. 1985, 197, 4586.
(59) Arase, A.; Hoshi, M.; Mijin, A.; Nishi, K. Synth. Commun. 1995, 25, 1957.
(60) Hanessian, S.; Meng, Q.; Olivier, E. Tetrahedron Lett. 1994, 35, 5393.
(61) Paterson, I.; Watson, C.; Yeung, K.-S.; Wallace, P. A.; Ward, R. A. J.
Org. Chem. 1997, 62, 452.
(56) Macrocyclization substrate 33 was synthesized from intermediates generated
from the Horner and modified Julia coupling sequences.
(57) Attempts to cyclize seco acids 9 and 33 were performed using the following
conditions: trichlorobenzoyl chloride, Et3N, DMAP, benzene, reflux;
2-chloro-N-methylpyridinium iodide, Et3N, CH3CN or CH2Cl2; BOP-Cl,
Et3N, CH2Cl2 or toluene; DCC, DMAP; and DCC, DMAP‚HCl, CH2Cl2.
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 6985

A R T I C L E S
Scheidt et al.
ketone 8b, it was disappointing that a considerable number of
protecting group manipulations were required in order to
orchestrate carefully the functionality at C(7), C(15), and C(19)
to facilitate both the macrolactonization and the ultimate
deprotection of C(19)-OH prior to the oxidation step that
delivered methyl ketone 8b. The synthesis of vinylboronic acid
39, although highly stereoselective, required seven steps from
intermediate 28, which in turn was synthesized by a 10-step
sequence starting from methyl (2R,3R)-2-methyl-2-hydroxybu-
tyrate (44).33 While it was readily apparent that replacement of
the C(19)-DMPM ether by a C(19)-TES ether could be
accomplished at the beginning of the synthesis, thereby saving
two steps in the overall sequence, we also wanted to minimize
manipulations of C(15)-OH and especially to avoid the vinyl
to alkynyl conversion employed in the elaboration of 28 to 30.
We suspected that the latter two issues could be confronted
simultaneously if we used a γ-methoxy-substituted allenylmetal
reagent for the synthesis of these intermediates, rather than the
γ-methoxyallylchromium reagent employed in the synthesis of
28.33,64
catechol borane and catalytic 9-BBN59 then provided vinyl
boronic acid 39 in 65% overall yield from 34. The Suzuki cross
coupling of 39 (1.0 equiv) and vinyl iodide 12 (1.0 equiv) using
the TlOH modified conditions introduced by Kishi provided seco
ester 40 in 65% yield.54 The carboxylic ester of this intermediate
was hydrolyzed, and the resulting seco acid cyclized to lactone
41 (52% overall) by application of the Yamaguchi mixed
anhydride protocol.55 Because we hoped to avoid further
protecting group manipulations of C(7)-OH, and because it has
been reported that the C(21) hydroxyl of bafilomycin A1 can
be oxidized selectively in the presence of C(7)-OH,30 we
removed the C(19)-TES ether under acidic conditions and then
attempted to oxidize C(19)-OH of the resulting macrocycle (38)
selectively by using the Swern protocol (the conditions that had
been employed by Gatti in the selective oxidation noted
above).30 However, this two-step sequence provided the C(7)-
keto-C(19)-alcohol 42 as the major product, along with the
C(7),C(19)-diketone.
Accordingly, it was necessary to protect C(7)-OH before
unmasking C(19)-OH. This was accomplished by treatment of
41 with triethylsilyl triflate (TES-OTf) and 2,6-lutidine in CH2-
Cl2 to protect C(7)-OH, followed by selective hydrolysis of the
less hindered C(19)-OTES ether by exposure to trifluoroacetic
acid in THF-H2O at -5 °C.62 This two-step sequence provided
the C(19) alcohol 43 in 76% yield. Finally, oxidation of 43 using
the Dess-Martin reagent63 provided the targeted methyl ketone
8b in near quantitative yield (98%).
Second-Generation Synthesis of Vinylboronic Acid 39.
Although we had developed a successful synthesis of methyl
(62) Nelson, T. D.; Crouch, R. D. Synthesis 1996, 1031.
(63) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155.
9
6986 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Total Synthesis of (−)-Bafilomycin A
A R T I C L E S
1
The synthesis of aldehyde 49 with a C(19)-TES ether
protecting group was performed starting from the known ethyl
ester 44.65 Thus, protection of the free hydroxyl group of 44 as
a TES ether provided 45 in 95% yield. Controlled reduction of
the methyl ester by using DIBAL in CH2Cl2 at -78 °C then
provided aldehyde 46, which was used immediately in the
subsequent (E)-crotylboration reaction with (S,S)-15 and pro-
vided the expected 16,17-anti homopropargyl alcohol 47.44
However, protection of the C(17)-hydroxyl of this intermediate
was surprisingly difficult. When standard conditions were
employed for this step (e.g., TBS-Cl, imidazole, DMF or TBS-
OTf, CH2Cl2, i-Pr2NEt), the C(17,19)-bis-TBS ether corre-
sponding to 48 was obtained as the major product. However,
when the silylation of 47 was performed by using TBS-OTf
and i-Pr2NEt in DMF,66 the trans-silylation of the C(19)-OTES
ether was suppressed and the mono-TBS ether 48 was obtained
in good yield. Finally, two-step oxidative cleavage of the vinyl
group delivered aldehyde 49 in near-quantitative yield.
level of stereoselectivity is less than what we hoped to achieve
in the synthesis of 52. Moreover, scant information was available
to enable us to judge the potential of the racemic allenyl zinc
and titanium reagents of type 50 to react with good selectivity
with chiral aldehydes such as 49.74 Accordingly, we sought to
prepare an allenylstannane reagent 50, since we anticipated that
it might be possible to devise a route to the nonracemic reagent,
if needed to achieve good results in the reaction with 49.75
Because existing methods for synthesis of γ-alkoxy allenyl-
metal reagents are incompatible with R ) H in 50, we elected
to employ the TMS-substituted allenylmetal reagent 51 as a
surrogate. After examining a number of possibilities, we settled
on the choice of Met ) SnBuCl2 in 51.76 Key to our synthesis
of 51 is the facility with which propargylstannanes isomerize
to allenylstannanes under Lewis acidic conditions.71,77 Accord-
ingly, the R-methoxyprogargyl stannane 54 serves as a stable,
isolable precursor to 51, which is generated in situ by treatment
with BuSnCl3. Subsequent addition of a range of simple
aldehydes provided the targeted anti-diol monoethers 55 with
g97:3 selectivity.76
The reactions of aldehydes with allenylmetal reagents have
been extensively studied in recent years.67-69 Especially note-
worthy are Marshall’s elegant studies and synthetic applications
of chiral allenyltin reagents, with which homopropargyl alcohols
are obtained with excellent selectivity in reactions with chiral
aldehydes.67,68,70,71 Accordingly, we were interested in develop-
ing the reaction of chiral aldehyde 49 with a suitable alkoxy-
substituted allenylmetal reagent 50 as a means of synthesizing
anti-R-methoxy homopropargyl alcohol 52, the immediate
precursor to vinyl boronic acid 39. The literature reveals
examples of reactions of aldehydes with γ-alkoxyallenyl zinc
and R-alkoxypropargyl titanium reagents that proceed with
modest (e4:1)72 to good (7:1 to 19:1)73 selectivity for the
targeted anti-diol mono ether units, respectively. However, this
A literature report indicates that racemic 54 can be synthe-
sized by treatment of propargyl methyl ether 56 with t-BuLi
and then addition of Bu3SnCl.78 However we were not able to
obtain isomerically pure 54 by using this method; a mixture of
the allenylstannane and propargylstannane was obtained from
this procedure. Accordingly, we prepared 54 by transmetalation
of the intermediate progaryllithium species with zinc chloride.79
The resulting allenylzinc reagent was then treated with Bu3-
SnCl to give the R-methoxy propargylstannane 54 in 97% yield
and with only traces of the allenylstannane regioisomer.
However, attempts to prepare a highly enantioenriched version
of 54 thus far have eluded us.80
(64) Takai, K.; Nitta, K.; Utimoto, K. Tetrahedron Lett. 1988, 29, 5263.
(65) Fra¨ter, G.; Mu¨ller, U.; Gu¨nther, W. Tetrahedron 1984, 40, 1269.
(66) Boschelli, D.; Takemasa, T.; Nishitani, Y.; Masamune, S. Tetrahedron Lett.
1985, 26, 5239.
(74) A recent report by Normant indicated that a 48:33:19 mixture of
diastereomers is obtained in the reaction of an alkoxyallenylzinc reagent
with R-(tert-butyldimethylsilyloxy)phenyl acetaldehyde: Poisson, J.-F.;
Normant, F. H. Org. Lett. 2001, 3, 1889.
(67) Chemler, S. R.; Roush, W. R. In Modern Carbonyl Chemistry; Otera, J.,
Ed.; Wiley-VCH: Weinheim, 2000; p 403.
(75) A recent report by Normant indicted that nonracemic allenylzinc reagents
can be prepared by kinetic resolution: Poisson, J.-F.; Normant, F. H. J.
Am. Chem. Soc. 2001, 123, 4639.
(76) Savall, B. M.; Powell, N. A.; Roush, W. R. Org. Lett. 2001, 3, 3057.
(77) Guillerm, G.; Meganem, F.; LeQuan, M.; Brower, K. B. J. Organomet.
Chem. 1974, 67, 43.
(78) Anies, C.; Lallemand, J.-Y.; Pancrazi, A. Tetrahedron Lett. 1996, 37, 5519.
(79) Zweifel, G.; Hahn, G. J. Org. Chem. 1984, 49, 4565.
(80) Roush, W. R.; Savall, B. M., unpublished research results, 2001.
(68) Marshall, J. A. Chem. ReV. 1996, 96, 31.
(69) Yamamoto, H. In ComprehensiVe Organic Synthesis; Trost, B. M., Ed.;
Pergamon Press: Oxford, 1991; Vol. 2; p 81.
(70) Marshall, J. A.; Perkins, J. F.; Wolf, M. A. J. Org. Chem. 1995, 60, 5556.
(71) Marshall, J. A.; Yu, R. H.; Perkins, J. J. Org. Chem. 1995, 60, 5550.
(72) Mercier, F.; Epsztein, R.; Holand, S. Bull. Soc. Chem. Fr. 1972, 690.
(73) Ishaguro, M.; Ikeda, N.; Yamamoto, H. J. Org. Chem. 1982, 47, 2225.
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 6987

A R T I C L E S
Scheidt et al.
impressive and significant that we have achieved to date. For
example, use of the related aldehyde 5733 in reactions with 54
results in a maximum 6:1 selectivity for the desired isomer 58.
We presume that the intrinsic diastereofacial selectivity of the
chiral aldehyde contributes to the magnitude of the observed
diastereoselectivity of these kinetic resolution-diastereoselective
aldehyde addition processes, but additional studies must be
performed in order to probe this hypothesis more fully.
Even though we were unable to develop an enantioselective
synthesis of 54, it was apparent that an opportunity existed for
kinetic resolution of racemic 54 in the reaction with chiral
aldehyde 49:75,81 if the rates of reaction of the two enantiomers
of the derived allenylstannane 51 with 49 are sufficiently
distinct, it should be possible to obtain homopropargyl alcohol
53 with excellent diastereoselectivity. In the event, treatment76
of aldehyde 49 with 1.2 equiv of R-methoxyprogargyl stannane
54 in hexane at -40 to -45 °C provided 53 with 4:1
diastereoselectivity, but in only 33% yield. The selectivity
increased to 8:1 when 2.5 equiv of the propargylstannane reagent
was used, and impressively, 20:1 selectivity was achieved when
5 equiv of the propargylstannane reagent was employed. By
using the latter set of conditions, the targeted homopropargylic
alcohol 53 was obtained in 84% yield after chromatographic
purification. These data are strongly suggestive of a kinetic
resolution, with the (S)-enantiomer of the intermediate allenyl-
stannane 51 reacting with 49 at a faster rate than the (R)-
enantiomer. Finally, treatment of 53 with K2CO3 in MeOH at
23 °C (1 h), or better still with Amberlyst A-26 resin in MeOH,82
then provided the targeted terminal alkyne 52 in 89-93% yield.
Homopropargylic alcohol 52 prepared by this sequence was
identical to the same intermediate prepared by the selective
triethylsilylation of 34 described previously.
Completion of the Total Synthesis of Bafilomycin A1.
Completion of the total synthesis required that we perform a
stereoselective aldol reaction between an appropriately protected
2,3-anti-aldehyde (cf., 5) and methyl ketone 8b. In our previous
studies, we had optimized the aldol reactions of lithium enolates
of methyl ketones and had found that the best selectivity was
achieved when the 2,3-anti-aldehyde 5 was protected as a PMB
ether.27,28,33 We had also surmized that the chlorotitanium
enolate aldol technology83 would be better suited for use with
the highly functionalized methyl ketone 8, owing to the strict
requirements for chelating functionality at C(15) in the lithium
enolate aldol reactions of 6.27,28,33 However, our inability to
remove a DMPM protecting group at the stage of macrocycle
32 established that we would not be able to complete the
synthesis by using the PMB-protected aldehyde 5 for the final
aldol coupling with 8. Clearly, a different protecting group
strategy for the aldehyde fragment was required.
In an effort to discover a compatible â-hydroxy protecting
group for the aldehyde component, a series of aldol reactions
was performed using model methyl ketone 59 and variously
protected aldehydes 60. We chose to use the titanium enolate
because of the high stereoselectivity of the reaction of 59 and
aldehyde 5, as we have disclosed previously.27,28,33 Unfortu-
nately, the aldol reactions of the chlorotitanium enolate of 59
and aldehydes 60a-c provided only modest levels of diaste-
reoselectivity, favoring the desired stereoisomer 61. Of these
aldehydes, 60c with a TBS ether protecting group is the most
desirable for use in completing the synthesis. However, an
alternative tactic for coupling of 8 and the TBS-protected
aldehyde would be required if 60c were to be used in the
synthesis.
This synthesis of 52 constitutes a significant improvement
over the previous route to the same intermediate. This second-
generation synthesis of 52 ultimately allows the vinylboronic
acid 39 to be prepared in only nine steps from the â-hydroxy-
R-methyl butyrate 44, vs 17 steps for the first-generation
sequence (also starting from 44) via intermediate 28.33
Studies are currently in progress to define the full scope and
synthetic utility of the reactions of racemic 54 with chiral
aldehydes. We have observed kinetic resolutions with other
chiral aldehydes, but the results with 49 remain the most
Studies by Evans84 and Paterson85 and unpublished work from
our laboratory86 have demonstrated that 2,3-anti-â-hydroxy
(83) Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.; Urp´ı, F. J. Am. Chem. Soc.
1991, 113, 1047.
(84) Evans, D. A.; Dart, M. J.; Duffy, J. L.; Yang, M. G. J. Am. Chem. Soc.
1996, 118, 4322.
(81) Marshall, J. A.; Wang, X.-J. J. Org. Chem. 1990, 55, 6246.
(82) Bach, J.; Berenguer, R.; Garcia, J.; Loscertales, T.; Vilarrasa, J. J. Org.
Chem. 1996, 61, 9021.
(85) Paterson, I.; Cumming, J. G.; Smith, J. D.; Ward, R. A. Tetrahedron Lett.
1994, 35, 441.
(86) Yasue, K.; Roush, W. R., unpublished work, Indiana University, 1994.
9
6988 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002

Total Synthesis of (−)-Bafilomycin A
A R T I C L E S
1
obvious, therefore, that a very mild set of conditions would be
required for global deprotection of the penultimate bafilomycin
synthetic intermediate.
We utilized the model aldol 61c to probe the conditions
necessary to remove the C(17)- and C(23)-TBS protecting
groups. Exposure of 61c to HF-pyridine in THF did not effect
any reaction after 24 h at ambient temperature. However,
treatment of 61c with either 1.0 M TBAF in THF or tris-
(dimethylamino)sulfonium difluorosilicate (TAS-F) in DMF
generated the enone 63 in 50% yield.90,91 In an attempt to
moderate the basicity of the TAS-F reagent, aldol 61c was
treated with 5 equiv of TAS-F and 10 equiv of water in DMF.
Remarkably, this reaction provided hemiketal 64 in 75% yield.
Although the exact constitution of the reagent generated from
TAS-F and water remains to be determined, this combination
as well as TAS-F alone (in DMF, CH3CN or CH2Cl2) has been
used in our laboratories to remove silyl ether protecting groups
from a variety of acid- and base-sensitive molecules.92
aldehydes undergo highly Felkin selective Mukaiyama aldol
reactions.87,88 Accordingly, exposure of model methyl ketone
59 to a premixed solution of TMS-Cl and Et3N (1:1, v: v) in
CH2Cl2 at -78 °C followed by addition of LiHMDS generated
the TMS enol silane 62 in >99% yield.61 A -78 °C solution
of the crude silyl enol ether and freshly prepared aldehyde 60c33
was then treated with BF3‚Et2O to provide a >95:5 mixture of
aldol products favoring the desired C(21)-(R) isomer 61c in 78%
yield. The stereochemistry of the aldol product was assigned
by using the characteristic 1H NMR analysis of the ABX pattern
for the protons at C(20) and C(21).89
Armed with these insights, the bafilomycin A1 total synthesis
was completed as follows. Methyl ketone 8b was converted to
the corresponding TMS enol silane by treatment with LiHN-
(SiMe3)2 in THF at -78 °C in the presence of TMS-Cl and
Et3N. Treatment of a -78 °C solution of the enol silane and
aldehyde 60c in CH2Cl2 with BF3‚Et2O in the presence of 4 Å
molecular sieves for 45 min (0.3 M final concentration)
generated the penultimate bafilomycin precursor 65 with a
diastereoselectivity of >95:5 (72% yield; 85% based on
recovered methyl ketone 8b). Reactions performed at lower
concentrations required much longer reaction times and gave
diminished yields of aldol 65. Finally, removal of the C(17)-
and C(23)-TBS ethers and the C(7)-TES ether was accomplished
as planned by treatment of 65 with large excesses of TAS-F
and water in DMF, thereby providing synthetic (-)-bafilomycin
A1 (1) in 93% yield. It is of interest to note that the deprotection
of the C(17)- and C(23)-TBS ethers occurred in less than 0.5
h, but the hindered C(7)-OTES required 4-5 h for complete
deprotection. Synthetic (-)-bafilomycin A1 was identified by
Confident that the final carbon-carbon bond of the bafilo-
mycin A1 could be formed with excellent control over the C(21)-
aldol stereochemistry, we turned our attention to finding
conditions for the final deprotection sequence. It is known that
iso-bafilomycin A1 undergoes ring contraction when treated with
TBAF,60 and it is also known that hemiketals in the bafilomycin
series are sensitive even to mild acidic conditions, especially
when C(17)-OH is protected as a TBS or TES ether.26,33 There
was also concern that the hindered C(7)-OTES ether of the
advanced macrocycle might be difficult to deprotect.13,17 It was
1
comparison to a natural sample by H and 13C NMR spectros-
copy, IR, optical rotation, and TLC mobility in several solvent
systems.
(87) Gennari, C. In ComprehensiVe Organic Synthesis; Trost, B. M., Fleming,
I., Heathcock, C. H., Eds.; Pergamon Press: New York, 1991; pp 629
(Chapter 2.4).
(88) Mukaiyama, T.; Narasaka, K.; Banno, K. Chem. Lett. 1973, 1011.
(89) Roush, W. R.; Bannister, T. D.; Wendt, M. D.; VanNieuwehnze, M. S.;
Gustin, D. J.; Dilley, G. D.; Lane, G. D.; Scheidt, K. A.; Smith, W. J. J.
Org. Chem. 2002, in press.
(90) Middleton, W. J. In Organic Syntheses; Freeman, C. H., Ed.; Wiley: New
York, 1990; Vol. VII; p 528.
(91) Noyori, R.; Nishida, I.; Sakata, J.; Nishizawa, M. J. Am. Chem. Soc. 1980,
102, 1223.
(92) Scheidt, K. A.; Chen, H.; Follows, B. C.; Chemler, S. R.; Coffey, D. S.;
Roush, W. R. J. Org. Chem. 1998, 63, 6436.
9
J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002 6989

A R T I C L E S
Scheidt et al.
these reagents, the difficult C(6)-C(8) anti-anti dipropionate
of bafilomycin was constructed in high yield with 85:15
selectivity, and the R-methoxypropargylation of 49 with the new
reagent 54 set the C(14) and C(15) stereocenters with 20:1
selectivity. Furthermore, the use of the Suzuki cross-coupling
reaction allowed for the convergent union of two highly
functionalized intermediates. The methyl ketone aldol reaction
that was utilized as the last key bond formation proceeded in
high yield and excellent diastereoselectivity for the newly
formed C(21) alcohol. Use of TAS-F for the final deprotection
is also a noteworthy development. The synthetic strategy
outlined here should be applicable to the synthesis of bafilo-
mycin analogues not available from manipulations of the
naturally available material, as well as to syntheses of related
macrolide antibiotics (e.g., formamicin (3)).93
Acknowledgment. Support provided the National Institute
of General Medical Sciences (GM 38436) is gratefully acknowl-
edged.
Supporting Information Available: Complete experimental
details and summaries of efforts to synthesize (E,E)-diene
precursors to 9 via the Horner-Emmons and modified Julia
olefination strategies (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
Conclusion
This synthesis of bafilomycin A1 illustrates a practical
application of the asymmetric (E)-crotylboration and diastereo-
selective R-alkoxypropargylation technology developed in this
laboratory. Among the number of stereocenters established using
JA017885E
(93) Powell, N. A.; Roush, W. R. Org. Lett. 2001, 3, 453.
9
6990 J. AM. CHEM. SOC. VOL. 124, NO. 24, 2002