General approach to epipolythiodiketopiperazine alkaloids: Total synthesis of (+)-chaetocins A and C and (+)-12,12′-dideoxychetracin A

Article abstract of DOI:10.1021/ja106869s

A highly stereoselective and systematic strategy for the introduction of polysulfides in the synthesis of epipolythiodiketopiperazines is described. We report the first total synthesis of dimeric epitri- and epitetrathiodiketopiperazines.

Full text of DOI:10.1021/ja106869s

Published on Web 09/24/2010
General Approach to Epipolythiodiketopiperazine Alkaloids: Total Synthesis
of (+)-Chaetocins A and C and (+)-12,12′-Dideoxychetracin A
Justin Kim and Mohammad Movassaghi*
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
Received August 1, 2010; E-mail: movassag@mit.edu
Abstract: A highly stereoselective and systematic strategy for the
introduction of polysulfides in the synthesis of epipolythiodike-
topiperazines is described. We report the first total synthesis of
dimeric epitri- and epitetrathiodiketopiperazines.
Dimeric epipolythiodiketopiperazine alkaloids comprise the core
bioactive constituents of the highly cytotoxic extracts from fungi
of the Chaetomium genus.1,2 Studies have recognized the central
role of the polysulfide bridge in the virulence of this class of
metabolites,3c,d noting an increase in potency with a corresponding
increase in the degree of sulfuration among related isolates.2 While
total syntheses of both monomeric4,5 and dimeric6,7 epidithiodike-
topiperazine alkaloids have been addressed in previous reports, late-
stage selective and efficient access to higher-order polysulfides
continues to remain a challenge.8 In 2009, we reported the first
total synthesis of the dimeric epidithiodiketopiperazine (+)-12,12′-
dideoxyverticillin A (1),6,9 and recently, Sodeoka reported the total
synthesis of (+)-chaetocin A (3)7,10 (Figure 1). To date, there have
been no reported total syntheses of dimeric epitri- and epitetrathiodike-
topiperazines. Herein, we present a general approach to dimeric
epipolythiodiketopiperazine alkaloids via syntheses of the di-, tri-,
and tetrasulfides (+)-chaetocin A (3), (+)-chaetocin C (5), and (+)-
12,12′-dideoxychetracin A (7), respectively.
Figure 1. Representative dimeric epipolythiodiketopiperazines.14
Scheme 1. Retrosynthetic Analysis
Complete stereochemical control in thiolation and precision in the
degree of sulfidation served as the central tenets of our retrosynthetic
strategy toward epipolythiodiketopiperazines, as delineated in Scheme
1. We envisioned that all sulfur homologues of epipolythiodiketopip-
erazines could be derived selectively and systematically from the
corresponding polysulfanes by final-stage intramolecular cyclization
onto a C15 iminium ion such as 10. Recognizing dithiol 11 as a key
intermediate, we strived to access all of the necessary polysulfanes by
oxidation and subsequent application of an advanced-stage stereo- and
regioselective C11 sulfuration of a common dimeric diketopiperazine
intermediate, 12. Such a solution would enable the controlled synthesis
of tetra-, hexa-, and octasulfurated congeners.
Py2AgMnO4 to provide our key intermediate (-)-18 in 55% yield.
The high chemo- and stereoselectivity of these transformations allowed
facile gram-scale access to advanced intermediates.
Our initial attempts to introduce two pairs of sulfur atoms in a
cis-selective fashion proved ineffective because of the intransigence
of the C15 and C15′ hydroxyls toward acid-promoted ionization
conditions. Both stereoelectronic and inductive factors of the serine
side chain adversely influenced the ionization potential of the
hemiaminals, and while solvent and field effects proved influential,
harsher ionization conditions were incompatible with nearly all of
our previously examined cis-thiolation strategies.15 Development
of milder conditions for the desired chemoselective and stereocon-
trolled thiolation chemistry proved to be imperative.16
The innate reactivity differences between the C11 and C15
hemiaminals permitted a regio- and stereoselective introduction of
sulfur critical to our general route to epidi-, epitri-, and epitetrathiodike-
topiperazines (Scheme 2). Exposure of tetraol (-)-18 to trifluoroacetic
acid in hydrogen sulfide-saturated nitromethane generated the sensitive
bisthiohemiaminal 20 in a highly diastereoselective fashion.17 Con-
centration of the reaction mixture followed by addition of isobutyryl
Our synthesis of dimeric diketopiperazine (-)-18, a common
precursor to target epipolythiodiketopiperazines 3-7, is illustrated in
Scheme 2. Bromocyclization of the diketopiperazine (-)-13,11a which
was synthesized in three steps from commercially available N-Boc-
L-tryptophan and L-serine methyl ester hydrochloride,12 afforded the
endo-tetracyclic bromide (+)-14 in 59% yield with high diastereose-
lectivity.13 Subsequent efforts to accomplish an N-methylation using
our previously developed conditions (K2CO3, MeI, acetone)11a suffered
from elimination of the ꢀ-siloxy group; however, kinetic deprotonation
followed by methylation at -40 °C effectively circumvented the
problem and afforded (+)-15 in 86% yield. A single-step exchange
of the silyl ether for an acetate then provided the key dimerization
substrate (+)-16 in 85% yield, and CoCl(PPh3)3-mediated reductive
radical dimerization provided the dimeric diketopiperazine (-)-17 in
49% yield.11 The dimer was selectively tetrahydroxylated6 using
9
14376 J. AM. CHEM. SOC. 2010, 132, 14376–14378
10.1021/ja106869s  2010 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2. Enantioselective Total Synthesis of Dimeric Epipolythiodiketopiperazines (+)-3-7a
a Conditions: (a) Br2, PhH, 59%. (b) LHMDS, MeI, DMPU-THF, -78 f -40 °C, 86%. (c) HF·py, THF, py; AcCl, 85%. (d) CoCl(PPh3)3, CH2Cl2,
49%. (e) Py2AgMnO4, CH2Cl2, 55%. (f) H2S, TFA, MeNO2; iPrCOCl, CH2Cl2, 53% (two steps). (g) hν (350 nm), L-ascorbic acid, 1,4-dimethoxynaphthalene,
H2O, MeCN, 51%. (h) N2H4, THF, 0 °C; NaH, Ph3CSCl, 90%. (i) BF3 ·OEt2, 2,6-di-tBu-4-Me-pyridine, Et3SiH, CH2Cl2, 82%. (j) Otera’s cat., MeOH, PhMe,
85 °C, 92%. (k) N2H4, THF, 0 °C; TrSSCl, NEt3, 86%. (l) N2H4, THF, 0 °C, 93%. (m) TrSSSCl, NEt3, 80%. (n) TFAA, 2,6-di-tBu-4-Me-pyridine, MeCN;
BF3 ·OEt2, 91%. (o) HCO2Ac; MeCN, BF3 ·OEt2, 60%. (p) Otera’s cat., MeOH, PhMe, 90 °C; N2H4, 95%. (q) Ac2O, CH2Cl2, 70%. (r) HCl, MeOH, 52%.
chloride18 afforded the octacyclic dithioisobutyrate (+)-21 in 53% yield
(two steps). The isobutyrates served the dual role of preventing
detrimental hemiaminal opening under polar protic conditions and
activating the C15 hemiaminal function in an anticipated mild
ionization.
The unique challenges of C11 ionization in the presence of the C2
aminal substructure in this system compelled the precedence of C11
thiolation over N1 desulfonylation, a sequence opposite that in our
previously established methodologies for synthesis of epidithiodike-
topiperazines.6 Nonetheless, we discovered that irradiating a solution
of (+)-21 with a black-light phosphor-coated lamp in the presence of
the photosensitizer 1,4-dimethoxynaphthalene and the terminal reduc-
tant L-ascorbic acid cleanly provided the desired diaminodithioisobu-
tyrate (+)-22 in 51% yield.19,20 Chemoselective hydrazinolysis of the
thioesters in the presence of the primary acetates provided dithiol (+)-
23, which served as a strategic point of divergence en route to the
various polysulfide homologues.
Formation and sulfenylation of dithiol (+)-23, which could be
performed in a single step, afforded the corresponding bis(triphenyl-
methanedisulfide) (+)-24 in 90% yield (X ) S, Scheme 2).21
Subsequent ionization of the isobutyrates and efficient cyclization with
concomitant loss of a triphenylmethyl cation was accomplished (82%
yield) to give (+)-chaetocin A diacetate (4), a derivative used in the
isolation and characterization1 of (+)-3. Synchronized loss of the
triphenylmethyl cation was essential in harnessing the utility of the
rapidly disproportionating polysulfanes, a central component of our
synthetic strategy. Mild methanolysis of the acetates using Otera’s
catalyst at 85 °C in toluene afforded the natural product (+)-chaetocin
A (3) in 92% yield as a white solid.10 All of the spectroscopic data
for (+)-3 and (+)-4 matched those reported in the literature.1
We proceeded to examine the use of our stereoselective thiolation
strategy in the synthesis of more highly sulfurated and challenging
epipolythiodiketopiperazines. Bis(triphenylmethanetrisulfide) (+)-26
was smoothly obtained in 86% yield via the common dithiol (+)-23
when chloro(triphenylmethane)disulfane (X ) SS, Scheme 2) was used
as the sulfenylating agent. Attempts to cyclize (+)-26 to form the
corresponding epitrithiodiketopiperazine using the protocol developed
for the bisdisulfide (see above) were unsuccessful, however. The lower
rate of cyclization for formation of a larger ring may have enabled
competing decomposition pathways involving active participation of
the N1 lone pair. Trifluoroacetylation of the amines imparted sufficient
stability to allow subsequent C15 iminium ion formation, affording
trithiodiketopiperazine (+)-28 in 91% yield. Methanolysis of the
acetates followed by in situ hydrazinolysis of the trifluoroacetamides
provided (+)-chaetocin C (5) in 95% yield, offering the first total
synthesis of a dimeric epitrithiodiketopiperazine alkaloid. (+)-Cha-
etocin C (5) was converted to (+)-chaetocin C diacetate (6) for direct
9
J. AM. CHEM. SOC. VOL. 132, NO. 41, 2010 14377

C O M M U N I C A T I O N S
comparison with data reported in the isolation report.2 All of the
spectroscopic data for (+)-6 matched those reported in the literature.2,22
Interestingly, alkaloid (+)-5, which contains two epitrisulfide
bridges, exists as a 1:2.4:5.5 mixture of three distinct conformers
in chloroform at room temperature, with the intermediate conformer
being of heterodimeric constitution.23 The distribution changes by
structure, however, as variable-temperature NMR studies carried out
on the more stable derivative (+)-28 demonstrated that its three
conformers exist in a 1:1.5:2.4 ratio in toluene-d8 at 90 °C,24 with the
major conformer lacking a C2 symmetry axis. An activation barrier of
∼23 kcal/mol was determined12 for their interconversion in 1,3,5-
trimethylbenzene-d12.25 We report the first complete set of character-
ization data for all three conformers of (+)-5 and (+)-6.12
Our general strategy for the stereocontrolled synthesis of epipoly-
thiodiketopiperazines proved effective even for the synthesis of
congeners possessing a tetrasulfide bridge, a motif found in natural
products such as (+)-chetracin A (8). The versatile dithiol intermediate
(+)-23 was converted to the bis(triphenylmethanetetrasulfide) (-)-29
in 80% yield using chloro(triphenylmethane)trisulfane (X ) SSS,
Scheme 2). In contrast to the disulfide series, cyclization of the resulting
tetrasulfide required conditions that were more favorable for the
reversible disengagement of the neighboring acetate from the acylimin-
ium ion. An adjusted window of reactivity was also observed for the
resulting product during postcyclization transformations. For example,
while N1-trifluoroacetylation of (-)-29 followed by ionization provided
the octasulfated homologue of 28, removal of the trifluoroacetamides
proved ineffective under a variety of nucleophilic and basic conditions.
After significant experimentation, we recognized that dimeric epitet-
rasulfides were marginally tolerant of acidic reaction conditions. Thus,
conversion of diamine (-)-29 to bisformamide 30 not only enabled
the successful synthesis of dimeric epitetrathiodiketopiperazine (+)-
31 but also allowed rapid removal of the acetyl and formyl groups by
acid-catalyzed methanolysis to provide (+)-12,12′-dideoxychetracin
A (7) in 52% yield. Unlike the epitrisulfides, the epitetrasulfide appears
as a single conformer in solution on the NMR time scale at room
temperature. The preferred helicity of the staggered sulfur atoms in
the solid state was identified through the X-ray structure solution of
(+)-7.12
References
(1) Von Hauser, D.; Weber, H. P.; Sigg, H. P. HelV. Chim. Acta 1970, 53,
1061.
(2) Saito, T.; Suzuki, Y.; Koyama, K.; Natori, S.; Iitaka, Y.; Knoshita, T. Chem.
Pharm. Bull. 1988, 36, 1942.
(3) (a) Greiner, D.; Bonaldi, T.; Eskeland, R.; Roemer, E.; Imhof, A. Nat. Chem.
Biol. 2005, 1, 143. (b) Cherrier, T.; Suzanne, S.; Redel, L.; Calao, M.;
Marban, C.; Samah, B.; Mukerjee, R.; Schwartz, C.; Gras, G.; Sawaya,
B. E.; Zeichner, S. L.; Aunis, D.; Van Lint, C.; Rohr, O. Oncogene 2009,
28, 3380. (c) Cook, K. M.; Hilton, S. T.; Mecinovic, J.; Motherwell, W. B.;
Figg, W. D.; Schofield, C. J. J. Biol. Chem. 2009, 39, 26831. (d) Isham,
C. R.; Tibodeau, J. D.; Jin, W.; Timm, M. M.; Bible, K. C. Blood 2010,
109, 2579. (e) Lakshmikuttyamma, A.; Scott, S. A.; DeCoteau, J. F.; Geyer,
C. R. Oncogene 2010, 29, 576.
(4) For the landmark synthesis of gliotoxin, see: Fukuyama, T.; Nakatsuka,
S.; Kishi, Y. Tetrahedron 1981, 37, 2045.
(5) For previous epidithiodiketopiperazine syntheses, see: (a) Overman, L. E.;
Sato, T. Org. Lett. 2007, 9, 5267, and references cited therein. (b) Hino,
T.; Sato, T. Tetrahedron Lett. 1971, 12, 3127. (c) Ottenheijm, H. C. J.;
Herscheid, J. D. M.; Kerkhoff, G. P. C.; Spande, T. F. J. Org. Chem. 1976,
41, 3433. (d) Williams, R. M.; Rastetter, W. H. J. Org. Chem. 1980, 45,
2625. (e) Wu, Z.; Williams, L. J.; Danishefsky, S. J. Angew. Chem., Int.
Ed. 2000, 39, 3866.
(6) Kim, J.; Ashenhurst, J. A.; Movassaghi, M. Science 2009, 324, 238.
(7) Iwasa, E.; Hamashima, Y.; Fujishiro, S.; Higuchi, E.; Ito, A.; Yoshida,
M.; Sodeoka, M. J. Am. Chem. Soc. 2010, 132, 4078.
(8) For prior epitrithiodiketopiperazine syntheses, see: (a) Brewer, D.; Rahman,
R.; Safe, S.; Taylor, A. Chem. Commun. 1968, 1571. (b) Murdock, K. C.;
Angier, R. B. Chem. Commun. 1970, 55. (c) Safe, S.; Taylor, A. J. Chem.
Soc. C 1970, 3, 432. (d) Reference 5b. (e) Ohler, E.; Tatruch, F.; Schmidt,
U. Chem. Ber. 1972, 105, 3658. (f) Reference 5c. (g) Curtis, P. J.;
Greatbanks, D.; Hesp, B.; Cameron, A. F.; Freer, A. A. J. Chem. Soc. Perkin
Trans. 1 1977, 180. (h) Coffen, D. L.; Katonak, D. A.; Nelson, N. R.;
Sancilio, F. D. J. Org. Chem. 1977, 42, 948. (i) Kirby, G. W.; Rao, G. V.;
Robins, D. J.; Stark, W. M. Tetrahedron Lett. 1986, 27, 5539. For prior
epitetrathiodiketopiperazine syntheses, see: (j) References 5b and 8b, c,
g-i. (k) Poisel, H.; Schmidt, U. Angew. Chem., Int. Ed. 1971, 10, 130. (l)
Yoshimura, J.; Sugiyama, Y.; Matsunari, K.; Nakamura, H. Bull. Chem.
Soc. Jpn. 1974, 47, 1215. (m) Strunz, G. M.; Kakushima, M.; Stillwell,
M. A. Can. J. Chem. 1975, 53, 295. (n) Kirby, G. W.; Robins, D. J.; Stark,
M. J. Chem. Soc. Chem. Commun. 1983, 812. (o) Waring, P.; Eichner,
R. D.; Tiwari-Palni, U.; Mullbacher, A. Aust. J. Chem. 1987, 40, 991.
(9) Son, B. W.; Jensen, P. R.; Kauffman, C. A.; Fenical, W. Nat. Prod. Res.
1999, 13, 213.
(10) Movassaghi, M. Presented at the 239th National Meeting of the American
Chemical Society, San Francisco, CA, March 2010; Paper ORGN 404.
(11) (a) Movassaghi, M.; Schmidt, M. A.; Ashenhurst, J. A. Angew. Chem.,
Int. Ed. 2008, 47, 1485. (b) Movassaghi, M.; Schmidt, M. A. Angew. Chem.,
Int. Ed. 2007, 46, 3725.
(12) See the Supporting Information for details.
(13) Diastereocontrol was derived from dipole minimization effects in low-
dielectric-constant media; the silyl ether function was imperative for the
solubility of the notoriously insoluble diketopiperazines.
(14) For the systematic positional numbering system used throughout this report,
We have developed a general approach to dimeric epipolythiodike-
topiperazines, with the solution yielding the first total synthesis of
dimeric epitri- and epitetrathiodiketopiperazines. The hallmark of our
synthesis of (+)-chaetocin A (3),10 (+)-chaetocin C (5), and (+)-
12,12′-dideoxychetracin A (7) is the high level of stereochemical
control and chemoselectivity in sulfidation of a common dimeric
diketopiperazine to access dimeric epipolythiodiketopiperazines, a
strategy consistent with a biogenetic hypothesis for divergent sulfidation
of a common diketopiperazine precursor.12 We have exploited our
understanding of the reactivity and structure of epipolythiodiketopip-
erazines to guide the development of the chemistry. The chemo- and
stereoselective nature of this synthesis offers facile access to com-
pounds with promising biological function.3
see p S3 in the Supporting Information.
(15) For discussions of our first- and second-generation solutions to dimeric
epidithiodiketopiperazines, see ref 6.
(16) For challenges in purification of a diastereomeric mixture of intermediates
in a route related to our first-generation solution to epidithiodiketopiperazine
synthesis, see pp S7-S14 and S27-S28 in the Supporting Information for
ref 7.
(17) Exceptional control in the thiolation was achieved by stereoinduction from
the proximal C3 stereocenter. Significant adverse C15 stereoinduction via
anchimeric stabilization of the acyliminium ion by the acetate was
minimized in high-dielectric-constant media.
(18) An acetate proved to be too labile at C11 under the subsequent desulfo-
nylation step, while attempts to use a pivaloate resulted in premature
ionization.
(19) Hamada, T.; Nishida, A.; Yonemitsu, O. J. Am. Chem. Soc. 1986, 108,
140.
(20) The thioesters proved to be stable under the reductive photochemical
conditions, in contrast to the corresponding thiols and polysulfides.
(21) For epidithiodiketopiperazine synthesis using sulfenyl chlorides such as
chloro(triphenylmethane)disulfane for thiol protection, see ref 5d. For
preparation of this reagent, see: Williams, C. R.; Britten, J. F.; Harpp, D. N.
J. Org. Chem. 1994, 59, 806.
Acknowledgment. We are grateful for generous financial
support by NIH-NIGMS (GM089732). M.M. is a Camille Dreyfus
Teacher-Scholar. J.K. acknowledges a National Defense Science
and Engineering Graduate Fellowship. We thank Amgen, Bristol-
Myers Squibb, and DuPont for additional financial support.
(22) The 1H NMR shift reported for one of the diastereotopic protons at C12 of
the major isomer is 3.60 ppm, but we found the resonance to be at 3.33
ppm.
(23) It is heterodimeric with respect to the conformation of the two trisulfides.
(24) The ratio was obtained at this temperature because trifluoroacetamide
rotamer-derived peak broadening at lower temperatures prevented accurate
integration of the peaks for the separate entities.
Supporting Information Available: Experimental procedures,
spectroscopic data, copies of 1H and 13C NMR spectra, crystal structure
of (+)-7 (CIF), and reassignments of several resonances for (+)-3, (+)-
4, and (+)-6. This material is available free of charge via the Internet
at http://pubs.acs.org.
(25) Low-dielectric-constant media were crucial for these studies, particularly
when heating to 160 °C for 1H NMR signal coalescence.
JA106869S
9
14378 J. AM. CHEM. SOC. VOL. 132, NO. 41, 2010