A convergent three-component total synthesis of the powerful immunosuppressant (-)-sanglifehrin A

Article abstract of DOI:10.1021/ja020091v

The potent immunosuppressive agent (-)-sanglifehrin A (5), initially discovered in a soil sample from Malawi, has been synthesized in a highly conver soil gent and stereocontrolled manner. The enantioselective approach relies on initial construction of th

Full text of DOI:10.1021/ja020091v

Published on Web 03/27/2002
A Convergent Three-Component Total Synthesis of the
Powerful Immunosuppressant (-)-Sanglifehrin A
Leo A. Paquette,* Maosheng Duan, Ingo Konetzki, and Christoph Kempmann
Contribution from the EVans Chemical Laboratories, The Ohio State UniVersity,
Columbus, Ohio 43210
Received January 17, 2002
Abstract: The potent immunosuppressive agent (-)-sanglifehrin A (5), initially discovered in a soil sample
from Malawi, has been synthesized in a highly convergent and stereocontrolled manner. The enantioselective
approach relies on initial construction of the iodovinyl carboxylic acid 14, which is coupled to tripeptide 59
in advance of a key macrolactonization step that generates 61a. An alternative protocol that involves the
linkage of 14 to 46 for possible construction of the large ring failed due to an inability to bring about a
corresponding macrolactamization maneuver. An efficient means for elaborating the C26-N42 spirolactam
western sector of 5 is also detailed. This requisite fragment was assembled through the proper adaptation
of consecutive aldol tactics for construction of the nine stereogenic centers, six of which are contiguous.
The first aldol process consisted of the tin triflate-mediated reaction of the aldehyde derived from 72 with
enantiopure ketone 73 to generate the syn C36-C37 relationship resident in 75. Once the conversion of
75 to 78 had been completed, the attachment to ketone 66 was effected with (+)-DIPCl, thereby setting
the C33-C34 relationship as anti. Once functional group modifications had given rise to 62, spirolactam-
ization was achieved to deliver predominantly 94, thereby setting the stage for the acquisition of vinyl
stannane 13 and its subsequent palladium-catalyzed Stille coupling to 61b. Controlled acidic hydrolysis
completed the synthesis of 5. Other important features of the present route are addressed where relevant.
prominent immunosuppressants are cyclosporin A (CsA, 1),9,10
FK506 (2),11 and rapamycin (3).12-18A more recent newcomer
Introduction
The immune system is a multicellular ensemble designed to
eliminate foreign entities from the body. The sophisticated
response brought into play when such an event occurs involves
the growth and proliferation of cells that recognize and
ultimately reject the substance.1-4 This phenomenon is triggered
as the result of signal transduction, that process wherein
extracellular molecules influence intracellular events.5-8 In the
past decade or so, several important signaling drugs have been
discovered that become intimately involved in the orchestration
of the immune response. These powerful biochemical tools
exhibit specific cellular effects that allow dissection of the
mechanisms of signal transduction at the molecular level, shed
light on intracellular signaling pathways involved in T-cell
activation, and make possible organ transplantation. The most
is the marine sponge metabolite pateamine A (4).19
While 1-3 mediate their immunosuppressive properties
through immunophilins, they do so in unique ways. Thus, 1
binds to cyclophilins, while 2 and 3 bind to protein receptors
known as FKBPs.20-27 Beyond this, calcineurin, the essential
enzyme involved in intracellular signal transduction emanating
(9) Sehgal, S. N.; Baker, H.; Eng, C. P.; Singh, K.; Ve´zina, C. J. Antibiot.
1983, 36, 351-354.
(10) Findlay, J. A.; Liu, J. S.; Burnell, D. J.; Nakashima, T. T. Can. J. Chem.
1982, 60, 2046-2047.
(11) Kino, T.; Hatanaka, H.; Hashimoto, M.; Nishiyama, M.; Goto, T.; Okuhara,
M.; Kohsaka, M.; Aoki, H.; Imanaka, H. J. Antibiot. 1987, 40, 1249-
1255.
(12) Ve´zina, C.; Kudelski, A.; Sehgal S. N. J. Antibiot. 1975, 28, 721-726.
(13) Sehgal, S. N.; Baker, H.; Ve´zina, C. J. Antibiot. 1975, 28, 727-732.
(14) Baker, H.; Sidorowicz, A.; Sehgal, S. N.; Ve´zina, C. J. Antibiot. 1978, 31,
539-545.
(15) Singh, K.; Sun, S.; Ve´zina, C. J. Antibiot. 1979, 32, 630-645.
(16) Swindells, D. C. N.; White, P. S.; Findlay, J. A. Can. J. Chem. 1978, 56,
2491-2492.
* To whom correspondence should be addressed. E-mail: paquette.l@
osu.edu.
(1) Thompson, A. W., Starzl, T. E., Eds. ImmunosuppressiVe Drugs: DeVelop-
ments in Anti-Rejection Therapy; Edward Arnold: London, 1994.
(2) Abbas, A. K.; Lichtman, A. H.; Pober, J. S. Cellular and Molecular
Immunology; Saunders: Philadelphia, 1991.
(3) Darnell, J.; Lodish, H.; Baltimore, D. Molecular Cell Biology, 2nd ed.;
Scientific American Books: New York, 1990.
(17) McAlpine, J. B.; Swanson, S. J.; Jackson, M.; Whittern, D. N. J. Antibiot.
1991, 44, 688-690.
(18) Findlay, J. A.; Radics, L. Can. J. Chem. 1980, 58, 579-590.
(19) Northcote, P. T.; Blunt, J. W.; Munro, M. H. G. Tetrahedron Lett. 1991,
32, 6411-6414.
(20) Siekierka, J. J.; Hung, S. H. Y.; Poe, M.; Lin, C. S.; Sigal, N. H. Nature
1989, 341, 755-757.
(4) Alberts, B.; Bray, D.; Lewis, J.; Raff, M.; Roberts, K.; Watson, J. D.
Molecular Biology of the Cell, 2nd ed.; Garland Publishing: New York,
1989.
(21) Harding, M. W.; Galat, A.; Uehling, D. E.; Schreiber, S. L. Nature 1989,
341, 758-760.
(5) Liu, J. Immunol. Today 1993, 14, 290-295.
(22) Rosen, M. K.; Standaert, R. F.; Galat, A.; Nakatsuka, M.; Schreiber, S. L.
Science 1990, 248, 863-866.
(6) Belshaw, P. J.; Meyer, S. D.; Johnson, D. D.; Romo, D.; Ikeda, Y.; Andrus,
M.; Alberg, D. G.; Schultz, L. W.; Clardy, J.; Schreiber, S. L. Synlett 1994,
381-391.
(23) Fretz, H.; Albers, M. W.; Galat, A.; Standaert, R. F.; Lane, W. S.; Burakoff,
S. J.; Bierer, B. E.; Schreiber, S. L. J. Am. Chem. Soc. 1991, 113, 1409-
1411.
(7) Schreiber, S. L. Science 1991, 251, 283-287.
(8) Schreiber, S. L.; Albers, M. W.; Brown, E. J. Acc. Chem. Res. 1993, 26,
412-420.
(24) Van Duyne, G. D.; Standaert, R. F.; Schreiber, S. L.; Clardy, J. J. Am.
Chem. Soc. 1991, 113, 7433-7434.
9
10.1021/ja020091v CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 4257-4270
4257

A R T I C L E S
Paquette et al.
from the T-cell receptor, becomes bound to the cyclophylin-
CsA and FKBP-FK506 complexes. In contrast, the FKBP-
rapamycin complex inhibits FRAP, a kinase involved in
interleukin-2 receptor-mediated T-cell proliferation.
Cyclosporin A, the fungal metabolite capable of blocking
activation of the entire immune system, is widely used for the
prevention of organ rejection in a variety of transplant surgeries.
Acting as a prodrug, it becomes active following entry into the
cytosol of T-cells and after formation of the cyclophilin-CsA
complex with resultant specific inhibition of the phosphatase
activity of calcineurin. Notwithstanding, 1 is not a perfect drug.
CsA induces severe toxicity to the kidney and central nervous
system, thereby preventing its use for the treatment of other
immune disorders such as rheumatoid arthritis. Because cal-
cineurin is responsible for both the immunosuppresive activity
and the toxicity of CsA, the hope that new analogues of 1 may
be less toxic is greatly diminished.
Background
The sanglifehrins possess an entirely unprecedented type of
structure having a macrocyclic lactone core containing the two
uncommon amino acids (S)-3-carboxy-piperazine and (S)-m-
tyrosine (m-hydroxyphenylalanine). Piperazic acids are not
unknown, but they occur in the monamycins,28,29 the antibiotic
L- 156,602,30-32 and the azinothricins33-36 exclusively acylated
at N2 rather than at N1. (S)-m-Tyrosine has found its widest
use in medicinal chemistry, most particularly to probe metabolic
pathways of the central nervous system.37 In addition, a
spirolactam subunit not previously encountered in another
natural product holds a prominent position in the western sector.
The sanglifehrins were discovered by a team of Novartis
scientists in the fermentation broths of Streptomyces flaVeolus
(A92-308110). Although early recognition was made of the
presence of a 22-membered macrocyclic ring in combination
with a [5.5]spirolactam subunit, endo- and exocyclic conjugated
E,E-diene domains, and an unusual peptidic backbone,38
a
Novel immunosuppressants with different mechanisms of
action are thus needed. From the standpoint of T-cell signal
transduction, CsA has revealed only one of the many proteins
that are involved in mediating the signaling events. New and
different immunosuppressants may help to unravel new protein
targets that play important roles in T-cell signal transduction.
For these reasons, the recently disclosed sanglifehrins 5-9
represent a particularly exciting family of immunosuppressive
natural products.
(28) Bevan, K.; Davies, J. S.; Hassall, C. H.; Morton, R. B.; Phillips, D. A. S.
J. Chem. Soc. C 1971, 514-522.
(29) Hassall, C. H.; Johnson, W. H.; Theobald, C. J. J. Chem. Soc., Perkin Trans.
1 1979, 1451-1458.
(30) Durette, P. L.; Baker, F.; Barker, P. L.; Boger, J.; Bondy, S. S.; Hammond,
M. L.; Lanza, T. J.; Pessolano, A. A.; Caldwell, C. G. Tetrahedron Lett.
1990, 31, 1237-1240.
(31) Kopka, I. E. Tetrahedron Lett. 1990, 31, 4711-4714.
(32) Caldwell, C. G.; Rupprecht, K. M.; Boudy, S. S.; Davis, A. A. J. Org.
Chem. 1990, 55, 2355-2361.
(33) Hale, K. J.; Delisser, V. M.; Manaviazar, S. Tetrahedron Lett. 1992, 33,
7613-7616.
(25) Bierer, B. E.; Somers, P. K.; Wandless, T. J.; Burakoff, S. J.; Schreiber, S.
L. Science 1990, 250, 556-559.
(34) Hale, K. J.; Delisser, V. M.; Yeh, L.-K.; Peak, S. A.; Manaviazar, S.; Bhatia,
G. S. Tetrahedron Lett. 1994, 35, 7685-7688.
(26) Michnik, S. W.; Rosen, M. K.; Wandless, T. J.; Karplus, M.; Schreiber, S.
L. Science 1991, 252, 836-839.
(35) Hale, K. J.; Cai, J.; Delisser, V. M. Tetrahedron Lett. 1996, 37, 9345-
9348.
(27) Van Duyne, G. D.; Standaert, R. F.; Karplus, P. A.; Schreiber, S. L. Science
1991, 252, 839-842.
(36) Hale, K. J.; Cai, J. Chem. Commun. 1997, 2319-2320.
(37) Bender, D. M.; Williams, R. M. J. Org. Chem. 1997, 62, 6690-6691.
9
4258 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Synthesis of Immunosuppressant (−)-Sanglifehrin A
A R T I C L E S
Figure 1. Retrosynthetic analysis of sanglifehrin A via two possible pathways.
defining X-ray crystallographic analysis was necessary to
establish unequivocally the absolute configurations at C17, C38,
and C40.39 This accomplishment also confirmed in precise
fashion the status of the four contiguous chiral centers spanning
C14-C17. This level of challenging structural novelty is
matched by the impressive immunosuppressant activity of
sanglifehrin A (5), the most abundant factor produced by this
microbial strain.40 The high affinity of 5 for cyclophilin A and
its capacity for inhibiting mitogen-induced B-cell proliferation
without influencing T-cell receptor-mediated cytokine produc-
tion are especially notable.41 Consequently, 5 exerts its effect
via a mode of action entirely different from that of CsA.
Although the precise underlying mechanism remains unknown,
hopes are high that our understanding of the immune response
at the molecular level will be significantly enhanced as these
matters become clarified.
have been reported by the Novartis44 and Paquette groups.45
Macrolide analogues of 5 have also been accessed.41 Herein
we describe in full the design and execution of a convergent
strategy for producing natural (-)-sanglifehrin A,46 the flex-
ibility of which can be expected to make rationally designed
congeners available as extension is warranted.
Initial Retrosynthetic Analysis. The Macrolactamization/
Macrolactonization Distinction
In planning the synthesis of 5, we focused our attention on
two potentially complementary pathways, one involving as-
sembly of the eastern sector by macrolactamization (route A),
the other consisting of a macrolactonization protocol (route B,
Figure 1). Both of these options call similarly for the construc-
tion and subsequent union of the vinyl stannane substituted
spirolactam 13 and iodovinyl carboxylic acid 14, and are
distinguished only on the basis of whether dipeptide 11 or
tripeptide 12 is generated early on. These considerations are
distinctively different than the other routes examined previously,
which are dependent on elaboration of the macrocyclic ring by
C19-C20 bond construction via intramolecular palladium-
mediated coupling42 or ring closing olefin metathesis.41,42c
Fragments 13 and 14 were designed such that the Stille process
could ultimately be applied to craft the central C25-C26 bond
of the exocyclic E,E-diene. In addition to primary concerns
The striking properties defined above have prompted signifi-
cant interest in the targeted preparation of sanglifehrin A. A
total synthesis was achieved by the Nicolaou research team in
1999,42 a relay synthesis was concomitantly devised by Met-
ternich et al.,43 and stereoselective routes to significant fragments
(38) (a) Fehr, T.; Oberer, L. Intern. Patent Appl. WO97/02285, January 23, 1997.
(b) Fehr, T.; Oberer, L.; Quesniaux, R. V.; Sanglier, J.-J.; Schuler, W.;
Sedrani, R. Intern. Patent. Appl. WO98/07743, March 3, 1998.
(39) Fehr, T.; Kallen, J.; Oberer, L.; Sanglier, J.-J.; Schilling, W. J. Antibiot.
1999, 52, 474-479.
(40) (a) Sanglier, J.-J.; Quesniaux, V.; Fehr, T.; Hofmann, H.; Mahnke, M.;
Memmert, K.; Schuler, W.; Zenke, G.; Gschwind, L.; Maurer, C.; Schilling,
W. J. Antibiot. 1999, 52, 466-473. (b) Zenke, G.; Strittmatter, U.; Fuchs,
S.; Quesniaux, V. F. L.; Sanglier, J.-J.; Fehr, T. J. Immunol. 2001, 166,
7165-7171.
(41) Cabrejas, L. M. M.; Rohrbach, S.; Wagner, D.; Kallen, J.; Zenke, G.;
Wagner, J. Angew. Chem., Int. Ed. 1999, 38, 2443-2446.
(42) (a) Nicolaou, K. C.; Ohshima, T.; Murphy, F.; Barluenga, S.; Xu, J.;
Winssinger, N. Chem. Commun. 1999, 809-810. (b) Nicolaou, K. C.; Xu,
J.; Murphy, F.; Barluenga, S.; Baudoin, O.; Wei, H.-X.; Gray, D. L. F.;
Ohshima, T. Angew. Chem., Int. Ed. 1999, 38, 2447-2451. (c) Nicolaou,
K. C.; Murphy, F.; Barluenga, S.; Ohshima, T.; Wei, H.; Xu, J.; Gray, D.
L. F.; Baudoin, O. J. Am. Chem. Soc. 2000, 122, 3830-3838.
(43) Metternich, R.; Denni, D.; Thai, B.; Sedrani, R. J. Org. Chem. 1999, 9632-
9639.
(44) (a) Ba¨nteli, R.; Brun, I.; Hall, P.; Metternich, R. Tetrahedron Lett. 1999,
40, 2109-2112. (b) Hall, P.; Brun, J.; Denni, D.; Metternich, R. Synlett
2000, 315-318. (c) Wagner, J.; Martin Cabrejas, L. M.; Grossmith, C. E.;
Papageorgiou, C.; Senia, F.; Wagner, J.; Nolan, S. P. J. Org. Chem. 2000,
65, 9255-9260.
(45) (a) Paquette, L. A.; Konetzki, I.; Duan, M. Tetrahedron Lett. 1999, 40,
7441-7444. (b) Duan, M.; Paquette, L. A. Tetrahedron Lett. 2000, 41,
3789-3792.
(46) Duan, M.; Paquette, L. A. Angew. Chem., Int. Ed. 2001, 40, 3632.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4259

A R T I C L E S
Paquette et al.
Scheme 1
Scheme 2
Figure 2. Retrosynthetic analysis of the C131-C25 sector.
regarding incorporation of protected piperazic acid subunits such
as 10 without racemization47,48 and efficient union of the major
subtargets, the plan called for paying strict attention to proper
installation of the numerous stereocenters. The spans across
C14-C17 and C30-C41 were clearly destined to be most
challenging. The stereochemical course of the spirolactam ring
closure also held interest and appeal.
Evaluation of an Alkyne Surrogate at the Chain
Terminus of 14
The initial thinking underlying incorporation of the sensitive
vinyl iodide functionality into 14 was to delay matters until as
late a stage as possible. Application of the Negishi carbo-
alumination reaction to 15 seemed appropriate to the task at
hand.49 In this context, the zirconocene dichloride- or diiodide-
promoted addition of trimethylaluminum to protected propargyl
alcohol 15 would be expected to proceed in cis fashion and
with high regioselectivity ultimately to deliver 14 after exposure
to elemental iodine (Figure 2). To address the feasibility of this
ploy, we elected to use a reagent-controlled aldol reaction
between 16 and 17 as the means to reach 15.
The preparation of 16 began with the ozonolytic conversion
of readily available 1850 to aldehyde 19 (Scheme 1). Wittig-
Horner chain homologation of 19 by reaction with methyl
4-(diethylphosphono)crotonate (20) in the presence of lithium
hexamethyldisilazide51,52 furnished principally the E,E-diene 21.
Removal of the other unwanted geometric isomers was easily
accomplished chromatographically, thus setting the stage for
reduction of the desired ester to the alcohol with diisobutyl-
aluminum hydride and subsequent perruthenate oxidation. This
two-step sequence routinely generated aldehyde 16 in very good
yield.
Access to reaction partner 17 was realized in the manner
shown in Scheme 2. Wacker oxidation of the enantiopure
oxazolidinone 2253 led to regiospecific introduction of a ketone
carbonyl group, which was subsequently converted into dithio-
ketal 24 in conventional fashion. For installation of the propionyl
substituent, the lithium enolate of 24 was allowed to react under
conditions of kinetic control with the acid chloride. This process
proceeded with high diastereoselectivity to deliver principally
17. As anticipated from the prior observations of Evans,54 the
stereogenicity of the newly generated chiral center in 17 was
not eroded during chromatographic purification as a consequence
of low kinetic acidity presumably brought on by allylic strain
effects.
(47) Ciufolini, M. A.; Xi, N. Chem. Soc. ReV. 1998, 27, 437-445.
(48) Depew, K. M.; Kamenecka, T. M.; Danishefsky, S. J. Tetrahedron Lett.
2000, 41, 289-292.
Conveniently, 17 was cleanly transformed into its (E)-boron
enolate on exposure to chlorodicyclohexylborane and ethyldi-
methylamine.55 Introduction of aldehyde 16 resulted in conver-
sion to the anti aldol 26 (Scheme 3). It is noteworthy that the
(49) (a) Negishi, E.; Van Horn, D. E.; King, A. O.; Okukado, N. Synthesis 1979,
501-502. (b) Yoshida, T.; Negishi, E. J. Am. Chem. Soc. 1981, 103, 4985-
4987. (c) Rand, C. L.; Van Horn, D. E.; Moore, M. W.; Negishi, E. J.
Org. Chem. 1981, 46, 4093-4096.
(50) The TBS analogue of 18 has been described previously: Smith, A. B., III;
Ott, G. R. J. Am. Chem. Soc. 1998, 120, 3935-3948.
(51) Roush, W. R. J. Am. Chem. Soc. 1980, 102, 1390-1404.
(52) Oppolzer, W.; Dupuis, D.; Poli, G.; Raynham, T. M.; Bernardinelli, G.
Tetrahedron Lett. 1988, 29, 5885-5888.
(53) Rivero, R. A.; Greenlee, W. J. Tetrahedron Lett. 1991, 32, 2453-2456.
(54) Evans, D. A.; Ennis, M. D.; Le, T.; Mandel, N.; Mandel, G. J. Am. Chem.
Soc. 1984, 106, 1154-1156.
9
4260 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Synthesis of Immunosuppressant (−)-Sanglifehrin A
A R T I C L E S
Scheme 3
The Second Generation Approach to 14
efficiency of this key step was improved if the workup
conditions defined by Paterson56 were applied. The fourth
contiguous stereogenic center was introduced as in 27 by direct
reduction of 26 with tetramethylammonium triacetoxyboro-
hydride.57
The dethioacetalization of 27 that was next planned met with
irksome problems when a variety of carbonyl unmasking
protocols were applied.58 Ultimately, we found the use of
phenyliodine(III) bis(trifluoroacetate)59 to be highly preferred
in this application, giving rise to 28 in 66% yield. Under these
conditions, the intramolecular ketalization occurred spontane-
ously.
We next set about to effect the conversion of 28 into 15, a
goal that was unexpectedly thwarted by two unfavorable
discoveries. Initially, the oxazolidinone exhibited substantial
reluctance to undergo oxidative cleavage.60 The efficiency with
which 28 is transformed into 29 rather than the carboxylic acid
could not be constructively diverted. Further, application of the
classical carboalumination conditions to 28 as well as numerous
variants thereof invariably proved too harsh for this substrate.
The terminal acetylenic functionality could not be suitably
modified even under conditions where steric congestion was
abated by removal of the silyl protecting group. We also had
occasion to consider an alternative synthetic tactic involving
prior removal of the chiral auxiliary that had served us so well
to this point. In the final analysis, these complications caused
us to install the vinyl iodide unit from the very outset.
The readily availability of aldehyde 3061 enabled us to test
the suitability of carrying a terminal vinyl iodide fragment
through the necessary homologation process (Scheme 4).
Coupling of 30 with the lithium salt of 20 proceeded unevent-
fully to install the dienyl ester functionality with a very
acceptable 9:1 preference for the (E,E)-isomer 31. Subsequent
reduction to the primary carbinol with diisobutylaluminum
hydride followed by oxidation with manganese dioxide proved
to be relatively efficient operations. Beyond that, the coupling
of 32 to the (E)-boron enolate of 17 and stereocontrolled hydride
reduction proceeded in a manner very closely paralleling our
earlier observations with the terminal alkyne. Once the dithio-
acetal group in 33 had been removed en route to 34, we were
particularly pleased to observe that reductive cleavage of the
oxazolidinone ring in this advanced intermediate could be
accomplished regioselectively on exposure to sodium borohy-
dride in aqueous THF as the reaction medium.62 Alternative
recourse to the more customarily utilized lithium borohydride
reagent in THF with methanol as a cosolvent resulted in
deiodination and was therefore not serviceable. We could now
attempt the two-stage oxidation of 35 to carboxylic acid 36. As
foreshadowed by successful model experiments, this transforma-
tion proceeded in notably efficient fashion (97%), as did
unmasking of the free secondary hydroxyl to generate 14. The
discovery that all of the compounds depicted in Scheme 4 are
relatively stable and readily handled was particularly satisfying.
Evaluation of the Macrolactamization Ring Closing
Protocol (Route A)
(55) Evans, D. A.; Ng, H. P.; Clark, J. S.; Rieger, D. L. Tetrahedron 1992, 48,
2127-2142.
(56) Paterson, I.; Yeung, K.-S.; Watson, C.; Ward, R. A.; Wallace, P. A.
Tetrahedron 1998, 54, 11935-11954. The specific experimental details
can be found on p 11946.
As indicated earlier, we envisioned one possible option for
proper elaboration of the complete C1-C25 sector of sanglifeh-
(57) Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am. Chem. Soc. 1988,
110, 3560-3578.
(58) Nakatsuka, M.; Ragan, J. A.; Sammakia, T.; Smith, D. B.; Uehling, D. E.;
Schreiber, S. L. J. Am. Chem. Soc. 1990, 112, 5583-5601.
(59) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 287-290.
(60) Evans, D. A.; Britton, T. C.; Ellman, J. H. Tetrahedron Lett. 1987, 28,
6141-6144.
(61) Nicolaou, K. C.; King, N. P.; Finlay, M. V. R.; He, Y.; Roschangar, F.;
Vourloumis, D.; Vallberg, H.; Sarabia, F.; Ninokovic, S.; Hepworth, D.
Bioorg. Med. Chem. 1999, 7, 665-697.
(62) Prashad, M.; Har, P.; Kim, H.-Y.; Repic, O. Tetrahedron Lett. 1998, 39,
7067-7070.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4261

A R T I C L E S
Paquette et al.
Figure 3. Retrosynthetic analysis of C1-C25 assembly by means of macrolactamization.
Scheme 4
Scheme 5
rin A to involve a pivotal macrolactamization step. From the
retrosynthetic perspective, the projected conversion of 38 into
37 demands judicious selection of the protecting groups defined
as R, R′, and R′′ (Figure 3). Similar considerations apply to the
step involving appendage of the dipeptide fragment. These
tactics are adequately precedented. Far more demanding of these
protective groups was the need to remove them once 38 was
accessed and to do so under conditions which were neither too
basic nor overly acidic. These restrictions surfaced because of
the sensitivity of the C1 carboxylate to alkaline hydrolysis, and
of the proclivity of the acetal spanning C15 and C17 to acid-
promoted cleavage.
With these considerations in mind, the 2-(trimethylsilyl)-
ethoxycarbonyl (Teoc) protecting group63 was installed on the
prove troublesome in view of their recognized usefulness in
peptide chemistry65 and in other settings involving highly
functionalized, structurally complex intermediates.66
The second paradigm, guided by a desire to deprotect the
carboxyl group concurrently with unmasking of the pair of
piperazic acid nitrogen atoms as in Scheme 5. Because large
supplies of the known 3964 were available to us, simple exchange
of the Boc substituents proved to be highly expedient. Our
expectations were that removal of the Teoc groups would not
(63) (a) Carpino, L. A.; Tsao, J. H.; Ringsdorf, H.; Fell, E.; Hettrich, G. J. Chem.
Soc., Chem. Commun. 1978, 358-359. (b) Campbell, A. L.; Pilipauskas,
D. R.; Khanna, I. K.; Rhodes, R. A. Tetrahedron Lett. 1987, 28, 2331-
2334.
(65) (a) Wu¨nsch, E.; Moroder, L.; Keller, O. Hoppe-Seyler’s Z. Physiol. Chem.
1981, 362, 1289-1292. (b) Rosowsky, A.; Wright, J. E. J. Org. Chem.
1983, 48, 1539-1541. (c) Shute, R. E.; Rich, D. H. Synthesis 1987, 346-
349.
(64) The enantiomer of 39 has been reported earlier; consult ref 33.
9
4262 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Synthesis of Immunosuppressant (−)-Sanglifehrin A
A R T I C L E S
Scheme 6
Scheme 7
amino nitrogens, led us to consider the esterification of 44 with
trimethylsilylethanol (TMSEOH).67 This option was exercised
following the acquisition of the functionalized m-tyrosine 44
by a modification of the Bender and Williams method37
involving the alkylation of commercially available oxazinone
42 (>98% ee) with m-(benzyloxy-carbonyl)benzyl bromide68
(43, Scheme 6). These steps were followed by the hydrogenoly-
sis of 45 over 10% Pd/C to free the NH2 group, and subsequent
peptidic coupling to Cbz-L-valine as promoted by O-benzo-
triazolyl-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HB-
TU).69,70 This pair of reactions proceeded quite smoothly, giving
rise to enantiopure 46 in 62% overall yield.
Reduction of intermediate 46 led to removal of the Cbz
substituent, thereby allowing for linkage to 14. Conditions
designed for HATU71 to provide the driving force for amide
bond formation resulted in proper integration of the entire C13-
C25 sector as in 47. The reaction between 41 and 47 necessary
to advance to 48 was most efficiently brought about in the
presence of N-ethyl-N′-(3-(dimethylamino)propyl)-carbodiimide
hydrochloride (EDC)67 and 4-pyrrolidinopyridine (4-PPy).72
At this juncture, removal of the Teoc and TMSE groups
became our next goal. Unfortunately, many attempts to liberate
49 met with failure. All of the conditions screened (e.g., TBAF,
TBAF/SiO2, HF, HF‚py, TASF, etc.) led to the formation of
uncharacterizable products. To gain insight into the source of
this complication, the methyl ester of piperazic acid was
prepared and stored for several hours at room temperature.
During this time, 50 (prepared under acidic conditions followed
by neutralization of the salt) was transformed in part into the
dehydro derivative 52 (perhaps via 51) as clearly recognized
by 1H and 13C NMR spectroscopy73 (Scheme 7). Polymerization
again materialized to a major extent. Careful examination of
the literature revealed that N2-acylated piperazic acids experi-
ence ready oxidation under a wide variety of conditions.47
(66) For example: (a) Meyers, A. I.; Roland, D. M.; Comins, D. L.; Henning,
R.; Fleming, M. P.; Shimizu, K. J. Am. Chem. Soc. 1979, 101, 4732-
4734. (b) Roush, W. R.; Coffey, D. S.; Madar, D. J. J. Am. Chem. Soc.
1997, 119, 11331-11332.
(67) Bourne, G. T.; Howell, D. C.; Pritchard, M. C. Tetrahedron 1991, 47, 4763-
4774.
(68) Obi, N.; Amada, J. Japanese Patent Appl. JP 98-126310.
(69) Dourtoglou, V.; Ziegler, J. C.; Gross, B. Tetrahedron Lett. 1978, 1269-
1272.
(70) Dourtoglou, V.; Gross, B.; Lambropoulou, Z. C. Synthesis 1984, 572-
(72) Hassner, A.; Krepski, L. R.; Alexanian, V. Tetrahedron 1978, 2069-2076.
(73) 1H NMR (300 MHz, CDCl3): δ 6.32 (br s, 1H), 3.78 (s, 3H), 3.25-3.20
(m, 2H), 2.45 (t, J ) 6.7 Hz, 2H), 1.90 (m, 2H). 13C NMR (75 MHz,
CDCl3): δ 165.5, 132.1, 51.9, 41.7, 21.0, 17.4.
574.
(71) O-(7-Azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluoro-phos-
phate: Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397-4398.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4263

A R T I C L E S
Paquette et al.
Scheme 8
Consequently, the inherent tendency of 49 and 50 to react in
this manner was not surprising. The rapidity of this process,
which was sufficient to deter effective direct macrolactamization,
even under extremely mild conditions was, however, not
anticipated.
tered for such large ring-forming steps.74 That the structural
assignment to this advanced intermediate is properly formulated
follows most convincingly from its electrospray high-resolution
mass spectrum, the clear delineation in the 500 MHz 1H NMR
spectrum of all the requisite resonances for its constituent
protons, and a fully compatible 13C NMR spectrum. The
phenolic hydroxyl was unmasked with trimethylsilyl triflate and
2,6-lutidine in CH2Cl2 solution.
Highly Diastereocontrolled Synthesis of the C1-C25
Domain via Macrolactonization (Route B)
With the demise of route A, the focus of our attention turned
to the interesting macrolactonization option that required access
to a tripeptide array patterned after 11. The ready availability
of 3933 allowed for its conversion via 53 to 54 on a multigram
scale according to the Hale procedure (Scheme 8). HBTU
proved to be a particularly effective agent in bringing about
the acylation of 54 at N1 with the m-tyrosine 55. The realization
that the phenolic hydroxyl in 55 requires no protection during
the amidation step is noteworthy. Conversion to Boc derivative
57 and hydrogenolysis were followed by union with Cbz-
protected L-valine to give tripeptide 58. Once the primary amino
group had been unmasked as in 59, the requisite coupling to
14 could be accomplished efficiently.
Elaboration of the C26-N42 Spirolactam Western Zone
Both disconnections envisioned at the outset (Figure 1) shared
in common the need for 13 or a protected form thereof. The
presence of an (E)-vinylstannane moiety and the high density
of stereogenic centers (with 11 of the 16 framework carbons
bearing stereochemical information) was certain to provide
challenging complexities. The critical role of this precursor
demanded the development of a route that would be amenable
to the reliable throughput of material. The retrosynthetic
prospects shown in Figure 4 are seen to be dependent on the
independent preparation of enantiopure building blocks 64, 65,
and 66, and their proper amalgamation via stereocontrolled aldol
This short series of steps provided 60 in straightforward
fashion. Our ultimate expectation of bringing about the chemose-
lective saponification of 60 at the methyl ester site and ring
closure to deliver synthetic 61a was indeed realized successfully
with an overall efficiency closely paralleling the norm encoun-
(74) Compare, for example: (a) Hanessian, S.; Ugolini, A.; Dube´, D.; Hodges,
P. J.; Andre´, C. J. Am. Chem. Soc. 1986, 108, 2776-2778. (b) White, J.
D.; Amedio, J. C. J. Org. Chem. 1989, 54, 736-738. (c) Parmee, E. R.;
Steel, P. G.; Thomas, E. J. J. Chem. Soc., Chem. Commun. 1989, 1250-
1252.
9
4264 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Synthesis of Immunosuppressant (−)-Sanglifehrin A
A R T I C L E S
Figure 4. Retrosynthetic analysis of C26-N42 construction via multiple aldol tactics.
Scheme 9
reactions. Our ultimate expectation for 62 was an intramolecular
cyclization that would set the aminal configuration at C37 with
high fidelity.
These considerations led us to consider the known75,76 (R)-
acyloxazolidinone 67 as a point of departure. By allylation of
the enolate of 67, it was possible to set the proper absolute
configuration R to the external carbonyl54 (Scheme 9). Subse-
quent conversion to the monoprotected 1,5-diol 69 was ac-
complished by sequential reductive cleavage of the chiral
auxiliary with LiBH4 in wet ether, transformation of the
relatively volatile alcohol to its TBDPS ether, and hydrobora-
tion-oxidation with 9-BBN. Following two-step oxidation to the
carboxylic acid, the latter was most efficaciously converted into
70 through adaptation of an activated anhydride protocol.77,78
Following highly enantioselective methylation of the enolate
anion of 70 to give 71, the chiral auxiliary was reductively
cleaved with lithium borohydride. This tactic allowed for the
(77) Chakraborty, T. K.; Suresh, V. R. Tetrahedron Lett. 1998, 39, 7775-7778.
(78) The formation of the N-pivaloyloxazolidinone as an insoluble byproduct
was noted if the same reaction was carried out under the conditions
described by Evans (Evans, D. A.; Gage, J. R.; Leighton, J. L. J. Am. Chem.
Soc. 1992, 114, 9434-9453).
(75) Go¨schke, R.; Cohen, N. C.; Wood, J. M.; Mailbaum, J. Bioorg. Med. Chem.
Lett. 1997, 7, 2735-2340.
(76) Thom, C.; Kocienski, P. Synthesis 1992, 582-586.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4265

A R T I C L E S
Paquette et al.
Scheme 10
hydroxyl as its triethylsilyl ether and follow that by regio-
selective reductive cleavage of 77 with diisobutylaluminum-
hydride.84 Careful Swern oxidation of the primary carbinol so
formed led to aldehyde 78.
The seven-carbon extension of 78 in the direction of C26-
C32 began with oxidation of 79 with pyridinium chlorochro-
mate,85 activation of the resulting carboxylic acid by coupling
to pivaloyl chloride, and condensation of the mixed anhydride
with (S)-4-(phenylmethyl)-1,3-oxazolidin-2-one (Scheme 11).
The elaboration of 80 via its monomethylation product 81 to
methyl ketone 66 provided no unforeseen difficulties and
proceeded with high reaction efficiencies. The conjoining of
66 with 78 was mediated by a second asymmetric aldol reaction,
this time involving the chiral Lewis-acidic (+)-B-chlorodiiso-
pinocampheylborane (DIP chloride) in the presence of triethyl-
amine. In view of the somewhat checkered history of this
reagent,86 we were compelled to verify the stereochemical
outcome in 83. To this end, we elected to transform 83 into 84,
and to analyze the resulting acetal 84 by NOE methods. As
shown in C, it was quickly made apparent that all four groups
positioned on the 1,3-dioxane ring were equatorially disposed.
Consequently, the stereogenicity established at C33 was as
programmed.
generation of alcohol 72, Swern oxidation of which smoothly
generated a pivotal aldehyde whose role it was to enter into
condensation with the tin(II) enolate of S-(73)79,80under substrate
control.81 In light of precedent, our expectation was that re-
face selectivity as in A would be significantly favored over the
si-face alternative B because of the differing nonbonded steric
interactions operative in these transition states.
We could now attempt the directed triacetoxyborohydride
reduction of 83 to give 84 in advance of incorporation of the
resulting trans diol into an acetonide ring as in 85. With this
intermediate in hand, it proved an easy matter to rid the molecule
of its three different silyl protecting groups in a single operation
with TBAF in THF. Following this treatment, diol 86 emerged
as a polar colorless oil in 92% yield.
We set as our next interim goal the keto amide 93. It was
hoped that 86 would initially be amenable to oxidative lacton-
ization. Tetra-n-propylammonium perruthenate (TPAP) failed
to give the desired product, resulting instead chiefly in
decomposition. Therefore, attention was directed to Swern
conditions. This protocol was less than optimal, giving rise to
lactol 87, a material highly sensitive to more advanced oxidative
conversion to the lactone. In fact, attempts in this direction
induced competitive elimination to afford 88 and fragmentation
leading to 89. Passage through transition states D and E is
therefore implicated.
The conditions shown, which required minimal screening to
uncover, gave rise to 74 in 92% yield as a chromatographically
separable 92:8 mixture of diastereomers. The stereochemical
assignment to this syn aldol product follows from a complete
COSY analysis performed on its OTBS/OBn analogue, in line
with previous successes realized upon application of the J-based
method to conformationally flexible systems (JH36,H37 ) 1.7
Hz).82 Hydroxyl-directed reduction of 74 with tetramethyl-
ammonium triacetoxyborohydride57 made available the anti diol
75, the two hydroxyls in which were differentiated by DDQ
oxidation under kinetic control83 (Scheme 10).
With formation of the terminal p-methoxyphenyl (PMP)
acetal,76 we could now attempt protection of the remaining
(79) Paterson, I.; Donghi, M.; Gerlach, K. Angew. Chem., Int. Ed. 2000, 39,
3315-3319.
(80) In the present study, 73 was prepared by conversion of commercially
available (S)-(+)-methyl 3-hydroxy-2-methylpropionate into its Weinreb
amide, followed by protection of the free hydroxyl as its PMB ether by
reaction with the 2,2,2-trichloroacetimidate under catalysis by triflic acid
[Paterson, I.; Nowak, T. Tetrahedron Lett. 1996, 37, 8243-8246]. The
subsequent addition of ethylmagnesium bromide furnished 73 in 91% yield.
(81) (a) Paterson, I.; Tillyer, R. D. Tetrahedron Lett. 1992, 33, 4233-4236. (b)
Paterson, I.; Norcross, R. D.; Ward, R. A.; Romea, P.; Lister, M. A. J.
Am. Chem. Soc. 1994, 116, 11287-11314.
(82) Murata, M.; Matsuoka, S.; Matsumori, N.; Paul, G. K.; Tachibana, K. J.
Am. Chem. Soc. 1999, 121, 870-871.
(83) Oikawa, Y.; Nishi, T.; Yonemitsu, O. Tetrahedron Lett. 1983, 24, 4037-
4040.
Fortunately, an alternative approach involving differential
protection and unmasking of the two hydroxyl groups was
9
4266 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Synthesis of Immunosuppressant (−)-Sanglifehrin A
A R T I C L E S
Scheme 11
feasible. From among many available options, the decision was
made to advance to the primary pivalate having triethylsilyl
bonded to the secondary carbinol as in 90 (Scheme 12). This
maneuver allowed for selective liberation of the OH group at
C41 by Dibal-H reduction and its conventional three-stage
conversion to the methyl ester 92. The amide moiety was
introduced by the treatment of 92 with freshly prepared
dimethylaluminum amide in CH2Cl2 at the reflux temperature.87
Subsequently, we took recourse to TBAF deprotection at C37
and oxidation of 62 at that site to provide 93 with exceptional
efficiency (Scheme 13).
full-matrix Newton-Raphson method to more accurately de-
termine the associated relative steric energy values. Under these
conditions, diastereomer F, which corresponds in stereochemical
detail to the western sector of sanglifehrin A, was found to be
10 kcal/mol more stable than G. The more expanded structures
H and I, which likewise differ only in the configuration at the
Considerable thought and computational assessment were
devoted to our evaluation of the possible stereochemical course
of the acid-catalyzed ring closure of 93 with generation of a
[5.5]spirolactam system. It will be recognized that all but two
carbons of the simplified structures F and G are stereogenic
centers. The Monte Carlo conformational searching method was
applied to F and G at the MM3 force field level. Each
conformational search was allowed 1000 iterations to find the
global minimum. The latter were further minimized using the
(84) Curtis, N. R.; Holmes, A. B.; Looney, M. G. Tetrahedron Lett. 1992, 33,
671-674.
(85) Spencer, R. W.; Tam, T. F.; Thomas, E.; Robinson, V. J.; Krantz, A. J.
Am. Chem. Soc. 1986, 108, 5589-5597.
(86) Paterson, I.; Goodman, J. M.; Lister, M. A.; Schumann, R. C.; McClure,
C. K.; Norcross, R. D. Tetrahedron 1990, 46, 4663-4684.
(87) Basha, A.; Lipton, M.; Weinreb, S. M. Tetrahedron Lett. 1977, 4171-
4174.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4267

A R T I C L E S
Paquette et al.
Scheme 12
Scheme 13
spirocenter, were also examined to allow hydrogen bonding to
play its usual role in guiding thermodynamic stability. Rel-
evantly, H is also favored, a position likely solidified to a yet
greater extent by the second intramolecular hydrogen bond it
is capable of forming.
On this basis, we had little doubt that the acid-catalyzed
spirolactamization of 93 would proceed to deliver predominantly
94 to gain the stabilization provided by the anomeric effects.
Indeed, 94 was isolated alongside its epimer C37 95 in a 7:1
ratio following the treatment of 93 with camphorsulfonic acid.
9
4268 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Synthesis of Immunosuppressant (−)-Sanglifehrin A
A R T I C L E S
Scheme 14
The further elaboration of 96 and subsequently 97 proved
straightforward, thereby setting the stage for conversion to the
(E)-vinyl stannane 13. The conditions preferred for this key step
involved preliminary conversion to bromoalkyne 97 by treatment
of 96 with N-bromosuccinimide in the presence of silver
nitrate,88 followed by exposure to tri-n-butylstannane, triphen-
ylphosphine, and [Pd2 (dba)3]‚CHCl3 in THF solution according
to the Guibe´ protocol.89
of p-toluenesulfonic acid (1 equiv) and boric acid (5 equiv) as
previously reported by a Novartis team.43 Except for the slightly
lower optical rotation of our synthetic sample ([R]22 -26.9
D
(c 0.06, CH3OH)) relative to that given for natural sanglifehrin
A ([R]20 -67.3 (c 0.99, CH3OH)),39 the spectroscopic and
D
chromatographic properties of 5 duplicated those exhibited by
the authentic substance.
To recapitulate, the total synthesis of (-)-sanglifehrin A has
been completed in 64 steps using a variety of bond construction
methods that have proven to be reliably stereocontrolled. The
involvement of a terminal iodovinyl group from the outset to
the penultimate stage of global construction is particularly
noteworthy. The methodologies used for much of the structural
assembly underscore the utility of chiral enolate based aldol
reactions of different type, the applicability of chelate-controlled
hydride delivery for the installation of anti 1,3-diol arrangements
in complex settings, and the meritorious nature of proper
functional group deployment. Last, the reliability of the large-
ring macrolactonization maneuver is seen to facilitate subunit
assembly and subsequent coupling, and as such warrant
consideration in the manner in which analogues of this class of
natural products are accessed.
Arrival at the Target
With the availability of both 13 and 61b, attention quickly
turned to their palladium-catalyzed coupling. Earlier studies by
our group that had targeted the synthesis of polycavernoside
A90 and analogues thereof91 had demonstrated the feasibility of
conjoining complex, highly functionalized subunits with
PdCl2(CH3CN)2 in DMF at room temperature. The adaptation
of these conditions in the present context resulted in smooth
operation of the Stille reaction to deliver sanglifehrin acetal 98
(Scheme 14). Ultimately, the difficult acidic hydrolysis of 98
to arrive at the title compound 5 was carried out in the presence
(88) Hofmeister, H.; Annen, K.; Laurent, H.; Weichert, R. Angew. Chem., Int.
Ed. Engl. 1984, 23, 727.
(89) Zhang, H. X.; Guibe´, F.; Balavoine, G. J. Org. Chem. 1990, 55, 1857.
(90) Paquette, L. A.; Barriault, L.; Pissarnitski, D.; Johnston, J. N. J. Am. Chem.
Soc. 2000, 122, 619.
Acknowledgment. We are grateful to Novartis for an
authentic sample of sanglifehrin and to John Hofferberth for
the MM3 calculations. Financial support was possible through
(91) Barriault, L.; Boulet, S. L.; Fujiwara, K.; Murai, A.; Paquette, L. A.; Yotsu-
Yamashita, M. Bioorg. Med. Chem. Lett. 1999, 9, 2069.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4269

A R T I C L E S
Paquette et al.
the acquisitoin of unrestricted funds from Eli Lilly and Co. M.D.
was the recipient of a Robert Mayer Graduate Fellowship, and
I.K. was a postdoctoral fellow funded by the Alexander von
Humboldt Foundation.
the course of the investigation reported here, including copies
of the 1H NMR spectra of 13, 14, 61b, and natural and synthetic
5 (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
Supporting Information Available: Complete experimental
procedures and spectral data for all compounds generated in
JA020091V
9
4270 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002