Total synthesis of bafilomycin A1 relying on iterative 1,2-induction in acyclic precursors

Article abstract of DOI:10.1021/ja011452u

The macrolide bafilomycin A₁ was synthesized starting from D-valine and D-mannitol as chiral progenitors of propionate units. Acyclic subunits corresponding to different parts of the molecule were constructed based on an iterative 1,2-asymmetric induction protocol as a distinctive feature of the synthesis. The assembly of two segments encompassing the entire carbon framework of the macrolide was achieved by using a Stille coupling. The resulting seco-ester was further manipulated to provide crystalline bafilomycin A₁ via a conventional carbodiimide-mediated Keck-type macrolactonization.

Full text of DOI:10.1021/ja011452u

10200
J. Am. Chem. Soc. 2001, 123, 10200-10206
Total Synthesis of Bafilomycin A1 Relying on Iterative 1,2-Induction
in Acyclic Precursors
Stephen Hanessian,* Jianguo Ma, and Wengui Wang
Contribution from the Department of Chemistry, UniVersite´ de Montre´al, C.P. 6128,
Succursale Centre-Ville, Montre´al, Que´bec H3C 3J7, Canada
ReceiVed June 13, 2001
Abstract: The macrolide bafilomycin A1 was synthesized starting from D-valine and D-mannitol as chiral
progenitors of propionate units. Acyclic subunits corresponding to different parts of the molecule were
constructed based on an iterative 1,2-asymmetric induction protocol as a distinctive feature of the synthesis.
The assembly of two segments encompassing the entire carbon framework of the macrolide was achieved by
using a Stille coupling. The resulting seco-ester was further manipulated to provide crystalline bafilomycin A1
via a conventional carbodiimide-mediated Keck-type macrolactonization.
Introduction
tory activity.7 On the other hand, a ring-expanded 18-membered
lactone analogue,8 iso-bafilomycin A1, in which the hemiacetal-
lactone carbonyl H-bond was present, was inactive possibly due
to the altered topology of the macrocycle.
The first total synthesis of bafilomycin A1 was reported by
Evans and Calter.9 Two other syntheses have been disclosed
since then by Toshima10 and Roush,11 and their respective
groups. The synthesis of segments of bafilomycin A1 has been
reported by Paterson,12 Roush,13 and Marshall14 and their
respective co-workers. Total syntheses of concanamycin F15
have been disclosed by Paterson16 and Toshima,17 respectively.
Yonemitsu and co-workers18 have also described the total
synthesis of hygrolidin.19
The bafilomycins, concanamycins, and hygrolidins are a small
subset of a family of 16-membered and 18-membered tetraenic
macrolactones respectively that belong to the hygrolide group
of macrolide antibiotics.1 Bafilomycin A1, 1 (Figure 1) was
isolated in 1983 from cultures of Streptomyces griseus sp.
sulfuru by Werner and Hagenmeier.2 It exhibited inhibitory
activity against G.positive bacteria and fungi.2 It has also shown
immunosuppressive activity,3 and selective potent inhibition of
vacuolar H+-ATPases,4 with potential applications in the
treatment of osteoporosis. The structure and absolute configu-
ration of bafilomycin A1, 1 were established by X-ray crystal-
lographic analysis,5 which confirmed the earlier assignments
made by Corey and Ponder1 based on NMR data and molecular
modeling. Inspection of the fine functional features in 1 reveals
a unique H-bonding network involving the lactone carbonyl,
the hemiacetal hydroxyl group, and an intervening C-17
hydroxyl group, which confer upon its three-dimensional
structure topological features that may have important biological
implications at the molecular level. Indeed, a single-crystal X-ray
analysis of a Grob fragmentation product6 in which the
macrolactone remained intact and the pseudosugar unit was
transformed into an ester maintained a H-bonding network
similar to 1 between the C17 hydroxyl group and the lactone
carbonyl. This product retained substantial H+-ATPase inhibi-
(7) (a) Sundquist, K.; Lakkakorpi, P.; Wallmar, B.; Va¨a¨na¨nen, K.
Biochem. Biophys. Res. Commun. 1990, 168, 309. (b) For SAR work on
bafilomycin A1 derivatives, see: Gagliardi, S.; Gatti, P. A.; Belfiore, P.;
Zocchetti, A.; Clarke, G. D.; Farina, C. J. Med. Chem. 1998, 41, 1883.
(8) Hanessian, S.; Meng, Q.; Olivier, E. Tetrahedron Lett. 1994, 35, 5393.
(9) Evans, D. A.; Calter, M. A. Tetrahedron Lett. 1993, 34, 6871.
(10) Toshima, K.; Jyojima, T.; Yamaguchi, H.; Noguchi, Y.; Yoshida,
T.; Murase, H.; Nakata, M.; Matsumura, S. J. Org. Chem. 1997, 62, 3271.
For preliminary communications, see: Toshima, K.; Jyojima, T.; Yamagu-
chi, H.; Murase, H.; Yoshida, T.; Matsumura, S.; Nakata, M. Tetrahedron
Lett. 1996, 37, 1069. Toshima, K.; Yamaguchi, H.; Jyojima, T.; Noguchi,
Y.; Nakata, M.; Matsumura, S. Tetrahedron Lett. 1996, 37, 1073.
(11) Scheidt, K. A.; Tasaka, A.; Bannister, T. D.; Wendt, M. D.; Roush,
W. R. Angew. Chem., Int. Ed. Engl. 1999, 38, 1652.
(12) Paterson, I.; Bower, S.; McLeod, D. M. Tetrahedron Lett. 1995,
36, 175.
(13) Roush, W. R.; Bannister, T. D. Tetrahedron Lett. 1992, 33, 3587.
Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993, 34,
8387.
(14) Marshall, J. A.; Adams, N. A. Org. Lett. 2000, 2, 2897. A total
synthesis of the seco-acid precursor to bafilomycin A2 has been completed
by the same authors: private communication, Prof. J. A. Marshall.
(15) (a) Woo, J.-T.; Shinohara, C.; Sakai, K.; Hasumi, K.; Endo, A. J.
Antibiot. 1992, 45, 1108. (b) Bindseil, K. U.; Zeeck, A. J. Org. Chem. 1993,
58, 5487.
(1) Corey, E. J.; Ponder, J. W. Tetrahedron Lett. 1984, 25, 4325.
(2) (a) Werner, G.; Hagenmaier, H.; Drautz, H.; Baumgartner, A.; Za¨hner,
H. J. Antibiot. 1984, 37, 110. (b) Werner, G.; Hagenmaier, H.; Albert, K.;
Kohlshorn, H.; Drautz, H. Tetrahedron Lett. 1983, 24, 5193.
(3) (a) Heinle, S.; Stuenkel, K.; Zaehner, H.; Drautz, H.; Bessler, W. G.
Chem. Abstr. 1989, 109, 142203. (b) Arzneim. Forsch. 1988, 38, 1130.
(4) Bowman, E. J.; Siebers, A.; Altendorf, K. Proc. Natl. Acad. Sci.
U.S.A. 1988, 85, 7972.For a review, see: Stone, D. K.; Crider, B. P.; Su¨dhof,
T. C.; Xie, X.-S. J. Bioenerg. Biomembr. 1989, 21, 605.
(5) (a) Everett, J. R.; Baker, G. H.; Dorgan, R. J. J. J. Chem. Soc., Perkin
Trans. 2 1990, 717. (b) Baker, G. H.; Brown, P. J.; Dorgan, R. J. J.; Everett,
J. R. J. Chem. Soc., Perkin Trans. 2 1989, 1073, 1940. (c) 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.
(6) Hanessian, S.; Tehim, A.; Meng, Q.; Granberg, K. Tetrahedron Lett.
1996, 37, 9001. For a related fragmentation induced by DAST, see: Gatti,
P. A.; Gagliardi, S.; Cerri, A.; Farina, C. Tetrahedron Lett. 1997, 38, 6949.
For examples in the carbohydrate series, see: (a) Binkley, R. W. J.
Carbohydr. Chem. 1990, 9, 771. (b) Zajac, W. W., Jr.; Dampawan, P.;
Ditrow, D. J. Org. Chem. 1986, 51, 2618 and references therein.
(16) Paterson, I.; Doughty, A.; McLeod, M. D.; Trieselmann, T. Angew.
Chem., Int. Ed. 2000, 39, 1308.
(17) (a) Toshima, K.; Jyojima, T.; Miyamoto, N.; Katohno, M.; Nakata,
M.; Matsumura, S. J. Org. Chem. 2001, 66, 1708. (b) Jyojima, T.; Katohno,
M.; Miyamoto, N.; Nakata, M.; Matsumura, J.; Toshima, K. Tetrahedron
Lett. 1998, 39, 6003. (c) Jyojima, T.; Miyamoto, N.; Katonho, M.; Nakata,
M.; Matsumura, S.; Toshima, K. Tetrahedron Lett. 1998, 39, 6007.
(18) (a) Makino, K.; Kimura, K.; Nakajima, N.; Hashimoto, S.; Yone-
mitsu, O. Tetrahedron Lett. 1996, 37, 9073. (b) Makino, K.; Nakajima, N.;
Hashimoto, S.; Yonemitsu, O. Tetrahedron Lett. 1996, 37, 9077.
(19) Seto, H.; Akao, H.; Furihata, K.; Otake, N. Tetrahedron Lett. 1982,
23, 2667.
10.1021/ja011452u CCC: $20.00 Published 2001 by the American Chemical Society
Published on Web 10/02/2001

Total Synthesis of Bafilomycin A1
J. Am. Chem. Soc., Vol. 123, No. 42, 2001 10201
Figure 1.
The structure of 1 presents a number of challenges that must
be considered as part of the synthesis plan, which we enumerate
as follows: (a) four propionate units must be generated with
high stereochemical control, (b) the linking of these subunits
to the tetraenic portion must be based on mild C-C coupling
methods, maintaining the integrity of stereogenic centers and
olefin geometries, (c) once assembled, the acyclic seco-acid must
be successfully converted to the macrolactone, (d) the hemiacetal
function in the pseudosugar unit must be protected from
â-elimination, and (e) orchestration of final steps in which the
judiciously selected protective groups should uneventfully
liberate the intact product. In their original paper, Werner and
Hagenmeier2 had alluded to the instability of bafilomycin A1
under even mildly acidic or basic conditions, thus heightening
the challenge of a total synthesis. The above considerations were
dealt with in often elegant ways in the previously reported
syntheses. Evans and Calter9 relied on aldol constructs for the
polypropionate segments, and a Stille coupling20 to generate
an intermediate macrocyclic lactone, which they engaged in a
final aldol coupling to create the pseudosugar unit en route to
1. Toshima and co-workers10 relied on a combination of
strategies, utilizing enantiopure building blocks such as (S)-3-
hydroxy 2-methyl propionic acid, ethyl (S)-lactate, sugars, and
aldol/Stille methodologies as in the Evans and Calter synthesis
to reach the target. Roush and co-workers11 capitalized on
crotylboration as a key step for propionate triad synthesis, and
assembled an intermediate macrolactone based on the Kishi
modification of a Suzuki-type coupling.21 The compatibility of
protective groups and their ultimate removal have loomed as
potential stumbling blocks at various stages in the previously
reported syntheses. For example, silyl ethers were extensively
used since their removal under mild fluoride-ion catalyzed
conditions as a final step was considered “safe”.22 However,
deprotection of silyl ethers at various positions proved to be
problematic, and the nature of the silyl protective group was
also critical.
Our disconnective analysis is shown in Figure 1 where the
target structure 1 would be generated by macrolactonization of
the seco-acid 2. This was envisaged to arise from vinyl stannane
3 and vinyl iodide 4 subunits via a Stille coupling20 followed
by protective group adjustments. Subunit 3 would arise from
propionate triads23 having the required absolute stereochemistry,
originating from D-valine 5 and D-glyceraldehyde 6 as chiral
nonracemic starting materials. Subunit 4 would also find its
stereochemical progeny in D-glyceraldehyde, thus capitalizing
on a common precursor approach that originates with D-
mannitol, 7. The asterisk and pound signs designating the
stereogenic centers in D-glyceraldehyde and D-valine respec-
tively can be traced back to individual fragments, and eventually
(21) (a) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457. (b) Uenishi,
J.-I.; Beau, J.-M.; Armstrong, R. W.; Kishi, Y. J. Am. Chem. Soc. 1987,
109, 4756.
(22) Scheidt, K. A.; Chen, H.; Follows, B. C.; Chemler, S. R.; Coffey,
D. S.; Roush, W. R. J. Org. Chem. 1998, 63, 6436.
(23) Hoffmann, R. W. Angew. Chem., Ind. Ed. 2000, 39, 2054. Hoffmann,
R. W. Angew. Chem., Ind. Ed. Engl. 1987, 26, 489.
(20) Stille, J. K.; Groh, B. L. J. Am. Chem. Soc. 1987, 109, 813.

10202 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
Hanessian et al.
Scheme 1a
a Conditions: (a) NaNO2, H2SO4, then CH2N2, 53%; (b) BOMCl, i-Pr2NEt, CH2Cl2, 85%; (c) Dibal-H, toluene, 87%; (d) Swern oxidation, then
Ph3PdCHCO2Me, CH2Cl2, 86%; (e) Me2CuLi, TMSCl, THF, -78 °C, 90%; (f) KHMDS, then Davis oxaziridine, THF, 80%; (g) H2, Pd(OH)2/C,
MeOH; (h) TBSOTf, 2,6-lutidine, CH2Cl2, 94%; (i) Dibal-H, toluene, 86%; (j) PivCl, Pyr., 88%; (k) H2, Pd(OH)2/C, MeOH; (l) TBSOTf, 2,6-
lutidine, CH2Cl2, overall 95% for two steps; (m) Dibal-H, toluene, 91%; (n) I2, imidazole, PPh3, toluene, 84%.
Scheme 2a
located at C-7 and C-23 respectively of bafilomycin A1 (Figure
1). Paramount to this strategy was the successful construction
of the required propionate triads in a highly stereocontrolled
manner. As an added challenge, we sought to achieve this
objective by utilizing a strategy that relies on a series of iterative
1,2-inductions24 in one- and two-directional25,26 protocols.
Scheme 1 summarizes the synthesis of the C21-C25 subunit
from D-valine. Conversion to the selectively protected γ-alkoxy-
R,â-unsaturated ester 9 based on standard methodology was
achieved in good overall yield. Conjugate addition of lithium
dimethylcuprate as in analogous cases27 led to the anti-adduct
10 as the major product in 90% yield. We have previously
shown that 1,2-induction in related cuprate additions proceeded
with very high stereoselectivity, regardless of the nature of the
terminal carbon atom (alkoxy,27 alkyl,24b or aryl28). Treatment
of the corresponding potassium enolate with the Davis oxaziri-
dine reagent29 afforded the anti/syn adduct 11 in excellent yield
and stereoselectivity. Hydrogenolysis led directly to the crystal-
line lactone 12, which was suitable for single-crystal X-ray
analysis, thus confirming the stereochemical outcome of the
cuprate and hydroxylation reactions arising from two consecu-
tive and highly stereocontrolled 1,2-inductions. Protection of
the hydroxyl group in 11 and functional group adjustments led
to the intended C21-C25 subunit as the primary iodide 15.
The elaboration of the C19-C24 subunit as a nucleophilic
dithian chiron is shown in Scheme 2. The readily available
γ-alkoxy-R,â-unsaturated ester 16 prepared from D-mannitol27
was subjected to a stereocontrolled cuprate addition to give 17
as the major product.24,27 Consecutive cycles of enolate
hydroxylation24a and cuprate addition afforded 18 harboring a
syn/anti/syn propionate triad with the desired stereochemistry
a Conditions: (a) ref 27; (b) ref 24a; (c) TBAF-HOAc/THF, 95%;
(d) Pd(OH)2/C, MeOH, H2; (e) NaIO4, MeOH-H2O; (f) 1,3-pro-
panedithol, BF3-Et2O, 83%; (g) NaBH4, THF-H2O, 95%; (h) TBDPSCl,
imidazole, THF, 88%; (i) 2, 2-dimethoxypropane, PPTS, CH2CI2, 95%.
at the R-hydroxy ester site. Stereochemical confirmation was
ascertained through X-ray analysis.24a It should be recalled that
the four new stereogenic centers comprising the equivalent of
two biosynthetic propionate units in 18 originate from a single
center in 16 by a series of stereocontrolled sequential 1,2-
inductions. We have previously commented on the remarkably
stereocontrolled consecutive 1,2-inductions in growing acyclic
chains comprising conjugate cuprate additions and enolate
hydroxylations.24,27 Possible transition states accounting for anti
and syn groups were suggested based on preferred trajectories
of approach, stereoelectronic factors, and 1,2-allylic strain. It
is difficult to speculate to what extent the local conformation
of the growing chain can influence the stereochemical outcome
of the conjugate addition reactions. X-ray crystal structures of
compounds harboring anti-C-methyl and hydroxyl combinations
as found in C5-C6 of 18 reveal an antiperiplanar orientation
on an extended carbon backbone. Insights into the conforma-
tional preferences of propionate triads have been recently
reviewed by Hoffmann.23 Conversion of 18 to the hemiacetal
19 was followed by treatment with 1,3-propanedithiol in the
presence of BF3-Et2O which afforded the lactone dithiane
(24) (a) Hanessian, S.; Wang, W.; Gai, Y.; Olivier, E. J. Am. Chem.
Soc. 1997, 119, 10034. (b) Hanessian, S.; Gai, Y.; Wang, W. Tetrahedron
Lett. 1996, 37, 7473.
(25) Hanessian, S.; Ma, J.; Wang, W. Tetrahedron Lett. 1999, 40, 4627,
4631.
(26) For examples of two-directional functionalization of acylic chains,
see: Poss, C. S.; Schreiber, S. L. Acc. Chem. Res. 1994, 27, 9. Still, W. C.;
Barrish J. J. Am. Chem. Soc. 1983, 105, 2487. For a review see: Magnuson,
S. R. Tetrahedron 1995, 51, 2167.
(27) Hanessian, S.; Sumi, K. Synthesis 1991, 1083.
(28) Hanessian, S.; Raghavan, S. Bioorg. Med. Chem. Lett. 1994, 4, 1697.
(29) (a) Davis, F. A.; Chen, B.-C. Chem. ReV. 1992, 92, 919. (b)
Vishwakayma, L. C.; Stringer, O. L.; Davis, F. A. Org. Synth. 1988, 66,
203. (c) Davis, F. A.; Stringer, O. D. J. Org. Chem. 1982, 47, 1774.

Total Synthesis of Bafilomycin A1
J. Am. Chem. Soc., Vol. 123, No. 42, 2001 10203
Scheme 3a
a Conditions: (a) 21, t-BuLi, THF-HMPA, -78 °C, then 15, 84%; (b) TBAF-HOAc, THF, 91%; (c) DMSO, Et3N, SO3‚Py, toluene, room
temperature, 89%; (d) ethynylmagnesium bromide, THF, -10 °C, 89%; (e) Dess-Martin, CH2CI2, 92%; (f) Super-Hydride, CH2CI2, -78 °C, 93%;
(g) t-BuOK, CH3I, THF, 86%; (h) MeOH, CSA, 86% based on 40% of recovered starting material; (i) TESCI, DMAP, THF-DMF (3:1), 89%; (j)
Bu3SnH, PdPh2CI2(cat.), THF, 87%.
intermediate 20 in excellent overall yield. Further manipulation
and functional group adjustment led to the desired C19-C24
chiron 21, ready to be coupled to the iodide 15.
cyclic hemiacetals. In the event that the C23 hydroxyl group
were also to become free, then the resulting dioxaspiro bis-
acetals would definitely hamper further progress toward the
synthesis.33 Thus, mild hydrolysis of 26 led to the corresponding
diol, which was selectively converted into the C15 TES ether
27 in excellent yield. Finally, treatment with tributyltin hydride
in the presence of a catalytic quantity of bistriphenylphosphine
palladium(II) chloride34 led to the desired vinyl stannane subunit
3 in a highly stereocontrolled manner starting from D-valine
and D-mannitol and proceeding via acyclic precursors.
The elaboration of the entire vinyl stannane subunit 3 (Figure
1) is shown in Scheme 3. Thus, generation of the dithian anion
from 21 with tert-butyllithium in a mixture of THF and HMPA
and condensation with the iodide 15 proceeded in 84% yield.
The installation of the vinyl stannane group required the
introduction of an acetylenic alcohol having an S-configuration
at C14. The TBDPS ether was selectively cleaved and the
resulting alcohol was oxidized to the aldehyde 23 with use of
the Doering-Parikh reagent,30 which proved to be the most
efficient. The mixture of acetylenic alcohols 24 and 25 could
be easily obtained by reaction of 23 with ethynylmagnesium
bromide. Oxidation of the readily separated undesired R-isomer
was done with use of the Dess-Martin reagent31 in excellent
yield, since Swern and Doering-Parikh oxidations were curi-
ously unsuccessful. On the other hand, oxidation with MnO2
resulted in an 80% yield of the corresponding ketone but the
reaction was slow. Reduction of the ketone with Super hydride32
led to the desired S-isomer, 25, thus allowing the use of the
two epimers of the acetylenic alcohols in a productive way.
Methylation of 25 with t-BuOK as base gave the ether 26. That
the oxidation/reduction sequence leading to 25 had indeed
proceeded with the desired stereochemical outcome would be
ascertained later through completion of the synthesis of 1. On
the basis of numerous model studies in this series, it was
necessary to remove the C15/C17 isopropylidene acetal at this
juncture and to use protective groups that would be removed
without affecting the final target structure or intermediates
leading to it. For example, cleavage of the acetal under acidic
conditions in the presence of a C19 carbonyl led invariably to
The construction of the vinyl iodide subunit 4 involved the
use of a common chiron 17 that was transformed to 28 through
a series of high-yielding steps (Scheme 4). Oxidation to the
aldehyde and chain extension led to the γ-alkoxy-R,â-unsatur-
ated ester 29. We were now in position to extend the potential
of these stereocontrolled conjugate cuprate additions and to test
their efficacy in a two-directional mode.25 Indeed, treatment of
29 with lithium dimethyl cuprate afforded a 90% yield of 30 in
which the requisite (and arduously accessible)25,35 anti/anti
orientation was secured. Conversion of 30 to the triol 31
followed by oxidative cleavage and treatment with the propy-
lidene phosphorane in refluxing benzene36 led, after protection,
to the trans-olefin 32, which was further transformed to the
pivalate ester 33 in high yield. It is of interest that an analogous
olefination reaction with diethyl 2-ethoxycarbonyl phosphono-
propionate in the presence of NaH in THF led to a 2:1 mixture
of inseparable trans/cis-R,â-unsaturated esters. To further
elaborate this chiron into 4, we needed to extend the chain at
both extremities. To that end, we chose to introduce the vinyl
iodide functionality first. The MOM protective group in 30 was
selectively removed with B-bromocatecheborane37 to afford 35,
(33) A spiroacetal was indeed isolated in the course of degradative studies
on bafilomycin A1; unpublished results, S. Hanessian and Q. Meng.
(34) Zhang, H. X.; Guibe´, F.; Balavoine, G. J. Org. Chem. 1990, 55,
1857.
(35) Marshall, J. A.; Perkins, J. F.; Wolf, M. A. J. Org. Chem. 1995,
60, 5556.
(30) Parikh, J. R.; von E. Doering, W. J. Am. Chem. Soc. 1967, 89, 5505.
(31) (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4156. (b)
Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277. For an
improved procedure, see: Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58,
2899.
(32) Faucher, A.-M.; Brochu, C.; Landry, S. R.; Duchesne, I.; Hantos,
S.; Roy, A.; Myles, A.; Legault, C. Tetrahedron Lett. 1998, 39, 8425.
(36) Lukas, J.; Ramakers-Blom, E. J.; Hewitt, G. J.; De Boer, J. J. J.
Organomet. Chem. 1972, 46, 167.

10204 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
Hanessian et al.
Scheme 4a
a Conditions: (a) Dibal-H, toluene, -78 °C, 89%; (b) MOMCI, Hunig’s base, CH2CI2, 92%; (c) TBAF, THF, 92%; (d) Swern oxidation, then
Ph3PdCHCO2Me, CH2Cl2, 87%; (e) Me2CuLi, TMSCI, THF, -78 °C, 90%; (f) KHMDS, Davis’ reagent, THF, 84%; (g) LiBH4, MeOH, 92%; (h)
H2, Pd(OH)2/C, MeOH, 90%; (i) NalO4, CH2Cl2(with 2% H2O), then Ph3PdC(CH3)CO2Et, benzene, reflux, 18 h, 85%; (j) TESCI, 2, 6-Lutidine,
THF, 95%; (k) Dibal-H, toluene, -78 °C, 89%; (l) PivCI, Pyr., CH2Cl2, 95%; (m) B-bromocatecholborane, CH2CI2, -78 °C, 30% for 35, 60% for
34; (n) TESOTf, 2,6-lutidine, CH2Cl2, 95%; (o) PPTS, CH2CI2-CH3OH (3:1), 93%; (p) Dess-Martin, CH2Cl2, 93%; (q) dimethyl (1-diazo-2-
oxopropyl)phosphonate, K2CO3, MeOH, 91%; (r) Cp2ZrCI2, Me3AI, H2O(cat.), -30 °C, then I2, 78%; (s) Dibal-H, toluene, -78 °C, 92%; (t)
MnO2, CH2Cl2, 90%; (u) methyl methoxyacetate, LiHMDS, THF, 86%; (v) MsCl, Pyr., DBU, 80%.
9,10
accompanied by a major quantity of diol 34 that was converted
to 35 in a two-step procedure (Scheme 4). Oxidation to the
aldehyde 36 under Dess-Martin conditions and chain-elongation
to the acetylenic intermediate 37 occurred in high yield,
especially with use of the diazophosphonate method38 (97%)
as compared to the venerable Corey-Fuchs procedure (72%).39
Iodination utilizing the versatile Negishi protocol40 followed by
deesterification with Dibal-H led to the allylic alcohol 38, which
was efficiently oxidized to the corresponding R,â-unsaturated
aldehyde 39 with MnO2. Initial attempts to introduce the vinylic
R-methoxy ester utilizing 2-methoxy trimethylphosphonoacetate
in the presence of different bases (NaH, NaHMDS, DBU/LiBr
in THF) led to equal amounts of cis- and trans-olefins which
could not be separated. The desired trans-dienic geometry was
achieved through a simple aldol condensation with lithium
2-methoxyacetate, followed by mesylation of the resulting
alcohol and â-elimination. Much to our satisfaction the resulting
subunit 4, obtained in 83% yield, was stereochemically pure
within the limits of our detection.41
related to 2 (Figure 1). Previous syntheses of bafilomycin A1
relied on a Stille cross-coupling protocol to append appropriate
subunits and produced segments which were further elaborated
en route to bafilomycin A1. While operationally similar, our
Stille cross-coupling presented the uncertainty of having larger
fully assembled reacting partners 3 and 4, with a potentially
problematic dithian group.42 A successful coupling to produce
the entire seco-ester of type 2 would still require the completion
of a number of unprecedented steps comprising selective
deprotections, marcrolactonization, and hemiacetal formation,
without undue side reactions (ex. â-elimination, dioxaspiro bis-
acetal formation, etc.). Preliminary reactions intended to survey
palladium reagents to effect the catalytic Stille cross-coupling
proved discouraging. Thus, no coupling or decomposition was
observed with PdCl2(MeCN)2, PdCl2(PPh3)2, or Pd(dppf)Cl2.
Utilizing PdCl2(PPh3)2 or Pd(dppf)Cl2 in the presence of Hunig’s
base afforded the desired product in 10% and 35% yields,
respectively. However, using (Pd(dppf)Cl2 and triphenylarsine43
provided the seco-ester 40 in 60% yield.
With chirons 3 and 4 in hand, we were now poised to engage
them in a Stille coupling20 to produce the protected seco-ester
At this juncture, it is of interest to comment on the cross-
coupling reaction. It is surprising that the coupling failed with
PdCl2(dppf) in DMF at 50 °C, conditions that were successful
in the case of smaller vinyl iodide and vinyl stannane partners.17a
Only after the addition of triphenylarsine43 and Hunig’s base44
did we succeed in achieving efficient coupling of subunits 3
and 4. Our plan to include the dithian group on the vinyl
(37) Boeckman, R. K., Jr.; Potenza, J. C. Tetrahedron Lett. 1985, 26,
1411.
(38) Ohira, S. Synth. Commun. 1989, 19, 561.
(39) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769.
(40) (a) Negishi, E.; Van Horn, D. E.; Yoshida, T. J. Am. Chem. Soc.
1985, 107, 6639. (b) Wipf, P.; Lim, S. Angew. Chem., Int. Ed. Engl. 1993,
32, 1068.
(41) The TBS ether variant of intermediate 4 has been synthesized by
Evans and Calter utilizing asymmetric aldol and Stille coupling methodol-
ogy: Calter, M. A. Ph.D. Dissertation, Harvard University, 1993; see ref
9.
(42) Smith, A. B., III; Condon, S. M.; McCauley, J. A.; Leazer, J. L.;
Leahy, J. W.; Maleczk, R. E., Jr. J. Am. Chem. Soc. 1997, 119, 962.
(43) Farina, V.; Krishnan, B. J. Am. Chem. Soc. 1991, 113, 9585.
(44) Smith, A. B., III; Ott, G. R. J. Am. Chem. Soc. 1996, 118, 13095.

Total Synthesis of Bafilomycin A1
J. Am. Chem. Soc., Vol. 123, No. 42, 2001 10205
Scheme 5a
a Conditions: (a) Pd(dppf)CI2, Hunig’s base, AsPh3, 50 °C, THF-DMF (1:1), 60%; (b) TBAF-AcOH (1:1), THF, 87%; (c) KOH, 80 °C, dioxane,
88%; (d) EDC, DMAP, reflux, CH2CI2, 65%; (e) TsOH, MeOH, 86%; (f) HgCl2, CaCO3, CH3CN-H2O (3:1), 85%.
stannane subunit 3 was fortuitous, since Smith and co-workers42
have shown that ligation of a proximal dithian group in a vinyl
iodide can affect the outcome of the cross-coupling reaction.
amounts of the 18-membered lactone (iso-bafilomycin A1).8 The
latter, shown to be energetically less favored by some 20-25
kcal/mol compared to bafilomycin A1,8 has been obtained
indirectly by ring expansion via an orthoacid intermediate in
the presence of an organocopper reagent, and its structure was
ascertained by X-ray crystallography.8 Fortunately, the EDC,
DMAP coupling protocol developed by Boden and Keck46
afforded the lactone 42 with no trace of iso-bafilomycin A1.
With the first of three major hurdles overcome, we proceeded
onward to the target molecule, with the daunting macrolacton-
ization in sight. To this end, we had to generate a partially
deprotected seco acid 41 from its precursor 40, which meant
selective cleavage of the C7 and C15 TES ethers in the presence
of the C21 and C23 TBS ethers, not to mention the base-induced
hydrolysis of the methyl ester to the carboxylic acid. In this
event, we found that a mixture of TBAF and acetic acid was
highly effective in the above-mentioned selective desilylations,
and the task was rendered even easier when basic hydrolysis of
the ester, even at 80 °C in aqueous dioxane, afforded the desired
seco-acid 41 in excellent overall yield.
There now remained to deprotect the two TBS ethers to afford
the penultimate intermediate 43, which was accomplished in
the presence of methanolic p-toluenesulfonic acid in 86% yield.
Finally, treatment of 43 with mercuric chloride and calcium
carbonate in aqueous acetonitrile effected smooth dethioacetal-
ization to afford crystalline bafilomycin A1, identical in all
respects with an authentic sample.
Previous syntheses9,10 of shorter subunits of 1 have succeeded
in the macrolactonization reaction adopting the Yamaguchi
protocol45 to give the desired 16-membered lactone. The
macrolactonization of a seco-acid consisting of the entire
bafilomycin A1 carbon framework in this series was unprec-
edented. Model studies in the macrolactonization of bafilomycin
A2 (the methyl glycoside) seco-acid utilizing EDC as described
by Boden and Keck46 led to a mixture of the desired lactone,
bafilomycin A1, and the corresponding glycal resulting from
elimination, in 30% yield. However, application of the same
protocol to the seco-acid derivative 41 led to the desired 16-
membered macrolactone 42 in 65% yield with no detectable
The fourth total synthesis of bafilomycin A1 reported in this
account highlights the power of stereochemical control in 1,2-
inductions in acyclic systems. In fact, a common chiral
progenitor with a single stereogenic center derived from
D-mannitol was the original “inducer”, eventually leading to four
consecutive enantiopure centers in the C15-C18 subunit. The
three contiguous stereogenic centers in the C6-C8 subunit were
generated from enantiopure R-2-hydroxy 3-methyl butyric acid,
easily prepared from R-valine. The installation of vicinal methyl
and hydroxy groups as part of four (or five) contiguous and
skipped propionate subunits with high stereochemical control
demonstrates the generality of the cuprate addition and enolate
hydroxylation protocol through a series of 1,2-inductions starting
with a single stereogenic center.24,25 Finally, the successful
application of the immensely useful Stille coupling, with highly
(45) Inanaga, I.; Hirata, K.; Saeki, T.; Katsuki, M.; Yamaguchi, M. Chem.
Bull. Soc. Jpn. 1979, 52, 1989.
(46) Boden, E. P.; Keck, G. E. J. Org. Chem. 1985, 50, 2394.

10206 J. Am. Chem. Soc., Vol. 123, No. 42, 2001
Hanessian et al.
functionalized units such as 3 and 4, further attests to the great
versatility of the method as a means of assembling complex
natural products through vinylic intermediates.
Dr. David Keeling (AstraZeneca) for biological tests of ana-
logues.
Supporting Information Available: Complete experimental
1
procedures, selected H, 13C spectra,, and an X-ray structure
Acknowledgment. We thank NSERCC and AstraZeneca
(Mo¨lndal, Sweden) for generous financial assistance through
the Medicinal Chemistry Chair Program. We thank Dr. Michel
Simard for X-ray structure determinations; we also thank
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA011452U