Stereocontrolled synthesis of trichodermatide A

Article abstract of DOI:10.1002/anie.201210099

Hard core made easy: The pentacyclic core of trichodermatide A was stereoselectively synthesized from a bis(1,3-cyclohexanedione) derivative by a ring-closing reaction followed by an intramolecular ketal formation (see scheme; PPTS=pyridinium p-toluenesulfonate). The first total synthesis of trichodermatide A was then completed by the introduction of three hydroxy groups. Copyright

Full text of DOI:10.1002/anie.201210099

.
Angewandte
Communications
DOI: 10.1002/anie.201210099
Natural Product Synthesis
Stereocontrolled Synthesis of Trichodermatide A**
Hiroki Shigehisa, Yoshihiro Suwa, Naho Furiya, Yuki Nakaya, Minoru Fukushima,
Yusuke Ichihashi, and Kou Hiroya*
The development of shorter synthetic sequences for the
synthesis of complex natural products represents significant
progress in synthetic organic chemistry. Considering this, the
development of a method capable of initially providing the
core structure of a target molecule followed by the stereo-,
regio-, and chemoselective introduction of functional groups
at the desired positions of the core skeleton is an efficient and
particularly attractive solution for the synthesis of complex
organic compounds.[1]
For the synthesis of the core of 1, we planned to use ketal
exchange and dehydrative pyran formation, both of which can
be promoted under acidic conditions, from dimerized 1,3-
cyclohexanedione derivative 5 in one pot (Scheme 1). Among
In 2008, Pei et al. elucidated the structures of trichoder-
matides A–D (1–4, Figure 1), which were isolated from the
marine-derived fungus Trichoderma ressei, and established
Scheme 1. Synthetic strategy for trichodermatide A (1).
Figure 1. Structures of trichodermatides A–D (1–4).
their cytotoxicity against the A375-S2 human melanoma cell
line.[2] Among them, trichodermatide A (1) is structurally
distinct from the other members of this family (2–4).
Trichodermatide A (1) incorporates a highly oxygenated
pentacyclic structure with eight chiral centers containing
a ketalic and hemiketalic carbon. Herein, we present the first
stereocontrolled total synthesis of 1 utilizing a method for the
efficient construction of core structure of 1 by a diastereose-
compounds 6–11, we expected that the synthetically potent
intermediates 10 or 11, both of which have stereogenic centers
corresponding to that of 1, would be selectively produced
under thermodynamic conditions. We have previously
reported a method for the construction of chiral quaternary
carbon centers on the basis of the diastereoselective synthesis
of the hemiketal and ketal derivatives of 2,2-disubstituted-1,3-
cyclohexanediones.[3]
lective ketal-formation reaction and
hydration reaction of enol ether.
a
cobalt-catalyzed
On the basis of this strategy, the synthesis commenced
with the preparation of aldehyde 12, which corresponds to the
side-chain portion of 1, from l-tartaric acid according to
known procedures.[4] Aldehyde 12 reacted with 1,3-cyclo-
hexanedione in the presence of piperidine in ethanol to
provide symmetric compound 5 in 89% yield (Scheme 2a).
The key synthesis step, which involved the diastereoselective
formation of a ketal from 5, was the initial attempt to reflux
benzene with p-toluenesulfonic acid (PTSA). Unfortunately,
however, the isolated compound was not one of the antici-
pated products 6–11, but was the unexpected bis(enol) ether
13, in which each hydroxy group of the side chain had reacted
with different carbonyl groups. In contrast, the reaction of 5
with a weak acid pyridinium p-toluenesulfonate (PPTS)
resulted only in the formation of a 9-substituted-tetrahydrox-
[*] Prof. Dr. H. Shigehisa, N. Furiya, Y. Nakaya, M. Fukushima,
Prof. Dr. K. Hiroya
Faculty of Pharmacy, Musashino University
1-1-20 Shinmachi Nishitokyo-shi Tokyo, 202-8585 (Japan)
E-mail: k_hiroya@musashino-u.ac.jp
Y. Suwa, Dr. Y. Ichihashi
Graduate School of Pharmaceutical Sciences, Tohoku University
6-3 Aramaki Aza Aoba, Aobaku, Sendai, 980-8578 (Japan)
[**] We gratefully acknowledge the useful suggestions of Prof. Dr.
Takayuki Doi (Tohoku University).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201210099.
3646
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3646 –3649

Angewandte
Chemie
the resulting acetonide was then hydrolyzed to give diol 15
(and its hemiketal mixture), which underwent a PPTS-
catalyzed ketal formation reaction to provide the potent
intermediate 11. The diastereoselective production of 11 is
presumably due to its greater thermodynamic stability
compared with the interconvertible diastereomer 9, because
the ketal ring exists in a chair form in 11 and a boat form in 9
(Scheme 2b). Given that the reaction of 11 with PTSA under
refluxing benzene gave 13 in 70% yield, the aforementioned
formation of 13 from 5 with PTSA would also occur through
compound 11 as a transient intermediate. The formation of
bis(enol) ether 13 from 11 apparently involved a 7-membered
ring cation intermediate 16, by energetically unfavorable
À
C O bond activation by PTSA. Therefore, PPTS is likely an
ideal reagent to avoid the acetal cleavage of 11.
The remaining tasks involved the introduction of three
hydroxy groups at the correct positions with the correct
stereochemistry. There are two ways to introduce the required
hydroxy groups onto 11. Of the available methods, we initially
decided to explore the introduction of the hydroxy group at
the C10 position by applying an allylic oxidation strategy
(Scheme 3).[8] The oxidation of 11 at the C10 position using
Scheme 2. a) Synthesis of the framework of trichodermatide A.
b) Mechanism and origin of diastereoselectivity. Hex=hexyl, PTSA=
p-toluenesulfonic acid, PPTS=pyridinium p-toluenesulfonate, AcOH=
acetic acid, n.O.e.=nuclear Overhauser effect.
anthene-1,8-dione ring, as represented by the conversion of 5
to 14.[5,6] Furthermore, the reaction of 14 under the same
conditions in the presence of PTSA also gave 13 in 60% yield.
These results suggested that the reaction of 5 with a strong
acid (PTSA) in the context of the ketal exchange reaction
produced the undesired product 13, and the tetrahydroxan-
thene-1,8-dione ring was cleaved under the same reaction
conditions. In this reaction, the participation of the secondary
hydroxy group(s) in cleaving the ring cannot be ruled out.[7]
Surprisingly, however, the pentacyclic compound 11 was
afforded in 74% yield as a single diastereomer together with
the undesired 13 (16% yield) by the reaction of 5 under weak
acidic conditions (acetic acid) in aqueous methanol followed
by treatment of the resulting intermediate with PPTS in
refluxing benzene. By careful analysis of the reaction, it
emerged that compound 5 was initially converted into 14 and
Scheme 3. Introduction of the hydroxy groups at C2 and C10.
DMAP=N,N-dimethyl-4-aminopyridine, KHMDS=potassium hexa-
methyldisilazide.
a stoichiometric quantity of SeO2 in benzene gave the desired
alcohol 17 in a stereoselective manner, although the yield was
only 15% with 11 being recovered in 20% yield. In spite of
our best efforts,[9] the yield of 17 could not be improved. The
diastereomer of 17 (C10-a-OH) was not detected because the
a face of the C8–C9 olefin is probably shielded by the axially
oriented C13-a-O ether bond. Although we could synthesize
17, the difficulty of further conversion forced us to alter the
route and investigate the introduction of hydroxy groups at
the C2 position first.
Angew. Chem. Int. Ed. 2013, 52, 3646 –3649
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3647

.
Angewandte
Communications
The reaction of the cross-conjugated potassium dienolate
of 11, which was kinetically generated with the Davis reagent
(18),[10] gave 20 as an inseparable diastereomeric mixture (a-
OH/b-OH = 4.9:1) in 46% yield. In anticipation of improve-
ments in diastereoselectivity, both enantiomers of the chiral
oxaziridine (À)-19 and (+)-19[11] reacted with the potassium
dienolate of 11. However, the selectivity decreased in the
reaction with (À)-19, whereas the major product of the
reaction with (+)-19 was b-OH; the origin of this diastereo-
selectivity is not clear. As the C10-oxidized compound could
not be afforded by the reaction of 20 with SeO2, the hydroxy
group was protected as the p-nitrobenzoate under standard
conditions (82% yield, inseparable diastereomeric mixture,
a-OR/b-OR = 4.9:1; R = p-nitrobenzoyl). The oxidation of 21
with SeO2 proceeded to provide 22 in 22% yield with
unreacted 21 recovered in 37% yield. A separation of the
diastereomers was carried out at this stage.
observed regioselectivity can account for the generation of
a C9 radical stabilized by the adjacent oxygen atom, which
would result from the cobalt–carbon bond cleavage in 23. The
C9 radical in 24 would subsequently react with molecular
oxygen to afford the peroxyradical 25, which would be then
converted into 26 by the cobalt catalyst and phenylsilane. The
stereochemistry of the C9 hydroxy group is fixed at the
b position by intramolecular hydrogen bonding with C10-b-
OH because of its hemiketal character. To the best of our
knowledge, this is the first example of the application of a Co-
catalyzed hydration reaction to an enol ether.
In summary, we have described the first synthesis of
trichodermatide A (1), which involves a short sequence
beginning with l-tartaric acid. The highly efficient intra-
molecular ketal formation reaction furnishes the pentacyclic
core of 1 by controlling two of the eight stereogenic centers.
Additional short sequences, including cobalt-catalyzed enol
ether hydration, successfully functionalize the core of 1 by
controlling the remaining four stereogenic centers. These
short step sequences should pave the way to a biological
development of 1 by efficiently synthesizing its analogues and
a molecule probe.
The introduction of the third hydroxy group by the
hydration of enol ether also proved problematic. In particular,
standard aqueous acidic conditions did not give any desired
products. After several efforts, we finally discovered the Co-
catalyzed hydration reaction developed by Isayama and
Mukaiyama (Scheme 4).[12,13] The reaction of 22a in the
Received: December 18, 2012
Published online: February 18, 2013
Keywords: cobalt · diastereoselective protection ·
.
natural products · polyketides · total synthesis
[1] a) N. Z. Burns, P. S. Baran, R. W. Hoffmann, Angew. Chem.
2009, 121, 2896 – 2910; Angew. Chem. Int. Ed. 2009, 48, 2854 –
2867; b) Y. Ishihara, P. S. Baran, Synlett 2010, 1733 – 1745.
[2] Y. Sun, L. Tian, J. Huang, H.-Y. Ma, Z. Zheng, A-L. Lv, K.
Yasukawa, Y.-H. Pei, Org. Lett. 2008, 10, 393 – 396.
[3] a) K. Hiroya, Y. Ichihashi, Y. Suwa, T. Ikai, K. Inamoto, T. Doi,
Tetrahedron Lett. 2010, 51, 3728 – 3731; b) K. Hiroya, Y. Suwa, Y.
Ichihashi, K. Inamoto, T. Doi, J. Org. Chem. 2011, 76, 4522 –
4532.
[4] a) G. Liu, Z. Wang, Synthesis 2001, 119 – 127; b) A. Fꢀrstner, M.
Wuchrer, Chem. Eur. J. 2006, 12, 76 – 89.
Scheme 4. Proposed mechanism for the conversion of 22a into 26
and for the total synthesis of trichodermatide A (1). acac=acetylaceto-
nate, R=p-nitrobenzoyl.
[5] For example, see: a) K. Rohr, R. Mahrwald, Bioorg. Med. Chem.
Lett. 2009, 19, 3949 – 3951; b) M. Bayat, H. Imanieh, S. H.
Hossieni, Chin. J. Chem. 2009, 27, 2203 – 2206; c) A. Sachar, P.
Gupta, S. Gupta, R. L. Sharma, Indian J. Chem. Sect. B 2009, 48,
1187 – 1194.
[6] For related recent examples for the cyclization reaction from
bis(1,3-cyclohexanedione) derivatives, see: a) G.-F. Han, B. Cui,
L.-Z. Chen, X. L. Hu, J. Heterocycl. Chem. 2012, 49, 195 – 199;
b) Y.-B. Shen, G.-W. Wang, Arkivoc 2008, xvi, 1 – 8; c) T.-S. Jin,
N. Qi, M. Li, L.-S. Han, L.-B. Liu, T.-S. Li, Asian J. Chem. 2007,
19, 3803 – 3809.
[7] The 9-Substituted-tetrahydroxanthene-1,8-dione ring was syn-
thesized in the presence of a strong acid with heating (Ref. [6a]:
H2SO4 in H2O, 70–808C, [6b]: MeSO3H, 1208C, and [6c]: p-
dodecylbenzenesulfonic acid in H2O, reflux).
[8] a) M. A. Umbreit, K. B. Sharpless, J. Am. Chem. Soc. 1977, 99,
5526 – 5528; b) M. Furber, L. N. Mander, J. Am. Chem. Soc.
1987, 109, 6389 – 6396.
presence of Co(acac)2 (acac = acetyl acetonate) and phenyl-
silane in trifluoroethanol under O2 atmosphere afforded 26
with satisfactory and chemo-, regio-, and stereoselectivity.
The reaction in THF, a commonly used solvent in cobalt
catalysis, barely hydrated the enol ether moiety in 22a. After
exchanging trifluoroethanol for methanol, p-nitrobenzoate
was hydrolyzed under basic conditions to afford trichoder-
matide A (1) in 53% yield from 22a. The spectral data of the
resulting synthetic 1 were compared with those reported in
the literature.[2]
The postulated origin of selectivity that is consistent with
experimental data and literature precedent is explained as
follows: By comparing the two olefins, the cobalt hydride
species likely reacts with nucleophilic enol ether (C8–C9)
more readily than the electron-deficient enone (C5–C6). The
axially oriented a ether oxygen on C13 could induce an
a facial approach of cobalt hydride by coordination; because
of this, C8 stereochemistry would be controlled.[14] The
[9] For example, we attempted to perform the reaction with
a catalytic amount of SeO2 in the presence of tert-butyl
hydroperoxide, with (PhSeO2)2O as an oxidizing agent, and
with SeO2 in a variety of different solvents (including dichloro-
3648 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 3646 –3649

Angewandte
Chemie
methane, 1,4-dioxane, acetonitrile, toluene, ethyl acetate, 2,2,2-
trifluoroethanol, and trifluorotoluene).
[10] L. C. Vishwakarma, O. D. Stringer, F. A. Davis, Org. Synth. Coll.
Vol. 1993, 8, 546 – 550, and references therein.
[11] a) J. C. Towson, M. C. Weismiller, G. S. Lal, A. C. Sheppard, A.
Kumar, F. A. Davis, Org. Synth. Coll. Vol. 1993, 8, 104 – 110, and
references therein; b) F. A. Davis, B.-C. Chen, Chem. Rev. 1992,
92, 919 – 934.
[13] For examples of Co-catalyzed hydration reactions in natural
products synthesis, see: a) J. Shi, G. Manolikakes, C.-H. Yeh,
C. A. Guerrero, R. A. Shenvi, H. Shigehisa, P. S. Baran, J. Am.
Chem. Soc. 2011, 133, 8014 – 8027; b) A. Kuramochi, H. Usuda,
K. Yamatsugu, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2005,
127, 14200 – 14201.
[14] T. Tokuyasu, S. Kunikawa, A. Masuyama, M. Nojima, Org. Lett.
2002, 4, 3595 – 3598.
[12] S. Isayama, T. Mukaiyama, Chem. Lett. 1989, 1071 – 1074.
Angew. Chem. Int. Ed. 2013, 52, 3646 –3649
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 3649