Synthesis of the entire carbon framework of the kedarcidin chromophore aglycon

Article abstract of DOI:10.1039/b705932a

In advanced studies directed toward the total synthesis of the kedarcidin chromophore, we have successfully achieved the late-stage installation of the nine-membered diyne ring in the presence of the highly functionalised ansamacrocyclic bridge. The Royal

Full text of DOI:10.1039/b705932a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Synthesis of the entire carbon framework of the kedarcidin
chromophore aglycon{
Fumihiko Yoshimura,a Martin J. Lear,*ab Isao Ohashi,a Yasuhito Koyamaa and Masahiro Hirama*a
Received (in Cambridge, UK) 19th April 2007, Accepted 20th June 2007
First published as an Advance Article on the web 29th June 2007
DOI: 10.1039/b705932a
In advanced studies directed toward the total synthesis of
the kedarcidin chromophore, we have successfully achieved the
late-stage installation of the nine-membered diyne ring in the
presence of the highly functionalised ansamacrocyclic bridge.
As highly unstable, complex natural products with unique
antiproliferative behaviours, the nine-membered chromophores
Scheme 1 Decomposition of chromophore model of kedarcidin through
of the enediyne chromoproteins are most worthy targets of
contemporary organic synthesis.1 A case in point is the kedarcidin
chromophore 1, which possesses an elaborate ansamacrocyclic
bridge and two unusual 2-deoxysugar components (Fig. 1).2
The major difficulty in synthesising the kedarcidin chromophore
1 pertains to surmounting the high enthalpic and entropic barriers
during the construction of the nine-membered, bicyclic core. This
difficulty is typically heightened by the short bench-lives of the
strained diyne products, whose CMC–C bond-angles are distorted
to around 160u. For example, besides the known spontaneous
Masamune–Bergman cycloaromatisations of fully-fledged epox-
yenediynes, diyne intermediates like 2 readily undergo Cope
rearrangements at ambient temperatures (Scheme 1).3 Clearly
then, the total synthesis of the kedarcidin chromophore demands
the late-stage installation of the nine-membered ring (and its
continued survival) within a complex chemical environment. In
their most recent study, Myers et al. overcame these barriers
through an elegant macrocyclic transannulation and have
a facile Cope rearrangement.
succeeded in the synthesis of 1, which has called into question
the natural stereochemistry at C10.4 In our approach, we have
centred on the established success of the CeCl3/LiN(TMS)2-
mediated cyclisation protocol.3 We report herein our advanced
progress toward the kedarcidin chromophore, which has resulted
in the synthesis of the multicyclic ansamacrolide (4).
In our retrosynthesis of 1, we elected to use the CeCl3/
LiN(TMS)2-mediated acetylide–aldehyde cyclisation reaction5
between C7 and C86 to form the nine-membered diyne core 4 of
the kedarcidin aglycon (Scheme 2). In addition to the concern of
the survival of such a complex, highly functionalised system, there
was the concern of the resulting stereochemical outcome at C8
(vide infra).
In order to supply sufficient quantities of the requisite precursor
(13) for cyclisation studies, efforts were initially directed at
improving our previous synthesis of the ansamacrolide 5.7 These
efforts have culminated in the total preparation of 3 g of the
ansamacrolide 5 by the development of efficient routes that have
enabled the large-scale production of all requisite fragments (see
supporting information for the practical synthesis of 5).
With gram quantities of the ansamacrolide 5 in hand, the
introduction of the naphthoamide unit to 5 was pursued
(Scheme 3). After a number of deprotection methods were
examined, the best chemoselective procedure to remove the Boc
Fig. 1 Structure of kedarcidin chromophore 1 (proposed 1997).2b
aDepartment of Chemistry, Graduate School of Science, Tohoku
University, Sendai 980-8578, Japan.
E-mail: hirama@mail.tains.tohoku.ac.jp; Fax: +81-22-795-6566
bDepartment of Chemistry, and Medicinal Chemistry Programme, 3
Science Drive 3, National University of Singapore, Republic of
Singapore 117543. E-mail: chmlmj@nus.edu.sg; Fax: +65-6779-1691
{ Electronic supplementary information (ESI) available: Experimental
procedures and characterization data for all new compounds. See DOI:
10.1039/b705932a
Scheme 2 Synthetic plans. PMBM 5 p-methoxybenzyloxymethyl.
Chem. Commun., 2007, 3057–3059 | 3057
This journal is ß The Royal Society of Chemistry 2007

View Article Online
Scheme 3 Reagents and conditions: a) TBSOTf, 2,6-lutidine, CH2Cl2, 250 to 0 uC; b) SiO2, CH2Cl2; c) 7, EDC?HCl, HOAt, CH2Cl2, 0 uC (89%, 3 steps);
d) allyl bromide, Cs2CO3, DMF, 0 uC (.99%); e) TBAF, THF, 0 uC (93%); f) TFA–THF–H2O (2 : 10 : 5), 50 uC (71%); g) TBSCl, Et3N, DMAP,
ClCH2CH2Cl (85%); h) SO3?pyridine, Et3N, CH2Cl2–DMSO (10 : 3) or IBX, MS4A, CH2Cl2–DMSO (10 : 3); i) TMSOTf, 2,6-lutidine, CH2Cl2, 270 uC.
Table 1 Nine-membered cyclisation study of 13 (Scheme 3)a
group in 5 first involved transformation to its corresponding
O-silylcarbamate, as described by Ohfune.8 This intermediate was
then exposed to silica gel prior to work-up procedures, thereby
ensuring the complete liberation of the free amine 6. Condensation
of 6 with the naphthoic acid 7 in the presence of EDC?HCl and
HOAt9 resulted in the smooth formation of its amide, which was
subsequently protected as its naphtholic allyl ether to afford a total
of 2.5 g of 9 in an 88% four-step yield from 5.
Entry
Additive
Temp. (uC)
Time (h)
Yield (%)
4a/4bb
1
2
3
4
5
6
a
CeCl3
CeCl3
CeCl3
CeCl3
none
225 to rt
215
225
250
225
1
18
25
69
36
36
,7
26
47
12
0
2/1
3/1
3/1
3/1
—
YbCl3
225
22
2/3
Reactions performed by adding 13 to 30–50 equivalents of pre-
mixed CeCl3 (or YbCl3)–LiN(TMS)2 [100 : 95] in THF [1 mM];
combined yields relate to 4a/b over 3 steps from 11, see Scheme 3.
At this stage, the selective transformation of the highly
oxygenated ansamacrolide 9 to a suitably protected cyclisation
precursor was not obvious to us, and the acetonide group in 9 was
found to be stubborn to deprotection. Eventually, we settled for
global desilylation with TBAF followed by hydrolysis of the
acetonide by using aqueous TFA at 50 uC to give the tetraol 10
(Scheme 3). Quite unexpectedly, treatment of 10 with TBSCl and
Et3N in the presence of a catalytic amount of DMAP gave the triol
11, which had been selectively protected at its C10 secondary
alcohol. This outcome was presumably due to the developing
transannular steric repulsions that stem from the ansamacrocyclic
bridge during the protection of the C8 primary alcohol.
b
Ratios determined on crude by 1H-NMR analysis (500 MHz).
entry 1). The reaction temperature affected both the yield and the
stereochemical outcome, and the best results were achieved at
225 uC to produce the cyclised products 4a and 4b in a 47%
combined yield, over three steps from 11 (Table 1, entries 2–4). In
contrast to the cyclisation of previous model systems, such as that
to afford 2,3 the cyclisation of 13 displayed an a-stereoselectivity at
the C-8 alcohol (cf. 4a), presumably due to the steric repulsion of
the ansamacrolide framework (cf. Scheme 4). Addition of CeCl3
The search for a reliable oxidation procedure to afford the
a-hydroxy aldehyde 12 proved to be an essential pre-requisite to
cyclisation studies. Oxidation of 11 under Dess–Martin or Swern
conditions gave complex mixtures, and simplified model systems
indicated that either oxidative cleavage of the diol group in 11 or
chlorination of the electron-rich naphthol unit occurred,10
respectively. Fortunately, IBX oxidation in the presence of
MS4A11 afforded 12. This development also led to the oxidation
of 11 by SO3?pyridine. Regardless of the method, the requisite
cyclisation precursor 13 was directly formed through the TMS-
O-silylation of 12, which was subsequently used in crude form.12
The ensuing cyclisation of 13 was first investigated under the
standard conditions that had been used to form 2,3 but the cyclised
products 4a and 4b were only isolated in low yield (Table 1,
Scheme 4 Attempted sulfonate formation of 4b. Reagents and conditions:
a) Dess–Martin periodinane, NaHCO3, CH2Cl2 (93%); b) Zn(BH4)2,
Et2O, 230 uC (57%).
3058 | Chem. Commun., 2007, 3057–3059
This journal is ß The Royal Society of Chemistry 2007

View Article Online
was indispensable for this reaction (Table 1, entry 5). Interestingly,
CeCl3 when replaced by YbCl3,13 reversed the stereoselectivity to
give 4b predominantly (Table 1, entry 6). Although 4a was not
stable at room temperature (half-life was approximately 13 h in
C6D6), 4a could be stored in a benzene matrix at 230 uC for one
week without deterioration. The C8-stereochemistry in 4a was
determined unambiguously by the NOE study of its mesylate (see
compound 18 in supporting information; definitive NOEs between
H8 and H49; H8 and H59). Moreover, the downfield-shift of key
13C-NMR data of 4a revealed that the alkyne carbons (C2/3, C6/7)
are under significant strain.
performed over 15 times to give 4a/b in yields of 38–47% (starting
from 30–90 mg of 11). In the meantime, we are addressing
several stereochemical issues4,19 and are striving to complete our
endeavours towards the total synthesis of the kedarcidin
chromophore.
This work was supported under the CREST program from JST.
A Fellowship to F.Y. from the Japanese Society for the Promotion
of Science for Young Scientists is gratefully acknowledged. The
authors wish to thank Dr Masako Ueno, Mr Kazuo Sasaki, Mr
Takeyoshi Kondo and Mr Toshio Sato (Research and Analysis
Center for Giant Molecules, Graduate School of Science, Tohoku
University) for mass spectral and NMR analysis.
In our earlier study, we succeeded in the epoxide formation
from b-alcohol at the C8-position of 2 via mesylation followed by
treatment with TBAF.3b The conversion of the alcohol at C8 of the
major cyclised product 4a to a suitable leaving group is thus key to
this transformation. Inversion of the secondary alcohol of 4a was
planned to be carried out via an oxidation–reduction sequence
(Scheme 4). After a number of investigations, 4a was oxidised to
the ynone 14 by using Dess–Martin periodinane in the presence of
NaHCO3. The chemoselective reduction of 14 was then examined.
The best reducing conditions involved freshly prepared Zn(BH4)2
in Et2O at 230 uC, producing the b-alcohol 4b in 57% yield. Other
reductants such as NaBH4–CeCl3?7H2O or LiB(sec-Bu)3H gave
complex mixtures, including the formation of TMS deprotected
products and over-reduced products.
Notes and references
1 Reviews: (a) J. W. Grissonm, G. U. Gunawardena, D. Klingberg and
D. Huang, Tetrahedron, 1996, 52, 6453; (b) S. J. Danishefsky and
M. D. Shair, J. Org. Chem., 1996, 61, 16.
2 Isolation and proposed structure of 1: (a) J. E. Leet, D. R. Schroeder,
D. R. Langley, K. L. Colson, S. Huang, K. E. Klohr, M. S. Lee,
J. Golik, S. J. Hofstead, T. W. Doyle and J. A. Matson, J. Am. Chem.
Soc., 1992, 114, 8432. Structure revision: (b) S. Kawata, A. Ashizawa
and M. Hirama, J. Am. Chem. Soc., 1997, 119, 12012.
3 (a) K. Iida and M. Hirama, J. Am. Chem. Soc., 1994, 116, 10310; (b)
S. Kawata, F. Yoshimura, J. Irie, H. Ehara and M. Hirama, Synlett,
1997, 250.
4 Synthesis of 1 and stereochemical revision of the kedarcidin chromo-
phore: F. Ren, P. C. Hogan, A. J. Anderson and A. G. Myers, J. Am.
Chem. Soc., in press.
5 Effect of CeCl3 on acetylide–aldehyde additions: (a) A. G. Myers,
P. M. Harrington and E. Y. Kuo, J. Am. Chem. Soc., 1991, 113, 694
(b) T. Nishikawa, M. Isobe and T. Goto, Synlett, 1991, 393.
6 The carbon numbering follows that for the kedarcidin chromophore
(1)2.
7 F. Yoshimura, S. Kawata and M. Hirama, Tetrahedron Lett., 1999, 40,
8281.
8 M. Sakaitani and Y. Ohfune, J. Org. Chem., 1990, 55, 870.
9 L. A. Carpino, J. Am. Chem. Soc., 1993, 115, 4397.
10 Aromatic-chlorinations with [Me2S+Cl]Cl2: G. A. Olah, L. Ohannesian
and M. Arvanaghi, Synthesis, 1986, 868.
11 M. Frigerio and M. Santagostino, Tetrahedron Lett., 1994, 35, 8019.
Addition of MS4A is required to suppress hydrate formation.
12 Compounds 12 and 13 are unstable to silica gel and should be used
immediately in the optimised three-step sequence from 13 to 4.
13 This reversion in stereoselectivity was anticipated on the basis of the
smaller ionic radius of Yb3+ as compared to Ce3+ (R. D. Shannon, Acta
Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Cryst., 1976, 32,
751) which would incur less steric interactions from the ansamacrocyclic
bridge during a nine-membered cyclisation, also see: K. Utimoto,
A. Nakamura and S. Matsubara, J. Am. Chem. Soc., 1990, 112, 8189;
D. Enders and J. Tiebes, Liebigs Ann. Chem., 1993, 173; S. Matsubara,
T. Ikeda, K. Oshima and K. Utimoto, Chem. Lett., 2001, 1226.
14 Y. Yoshida, K. Shimonishi, Y. Sakakura, S. Okada, N. Aso and
Y. Tanabe, Synthesis, 1999, 1633.
Having developed suitable conditions for the inversion of the
secondary alcohol of 4a, the transformation of the alcohol of 4b to
a suitable leaving group was investigated (Scheme 4). Mesylate
formation of 4b under various conditions, such as MsCl–Et3N,
MsCl–DMAP, MsCl–Me2N(CH2)6NMe2,14 or MsCl–AgO–KI15
did not give any mesylate 15.16 Next, triflate formation of 4b was
examined. However, triflate formation of 4b by using Tf2O17 in
the presence of various bases such as 2,6-di-t-butylpyridine, 2,6-
lutidine or pyridine, and TfCl–Et3N16 was unsuccessful.
Fluorosulfonate formation of 4b by using FSO2Cl,18 which would
act as a small sulfonation reagent, was also unsuccessful.
On the basis of these results, it is clear that our failure in forming
the mesylate, triflate, or fluorosulfonate forms of 4b is attributable
to steric repulsions of the ansa-bridge. An energy minimized model
structure (MM2* Macromodel ver. 6.0) of the simplified b-alcohol
compound showed that the alcohol at C8 is screened by the
ansa-bridge giving a small transannular cavity (see supporting
information). It seems, therefore, that an alternative synthetic
strategy is required for the construction of the nine-membered
epoxy diyne core of 1.
To conclude, we have succeeded in constructing the multicyclic
diyne ansamacrolide (4a/b), possessing the entire carbon skeleton
of the kedarcidin chromophore aglycon through the remarkable
facility of CeCl3 to moderate the anionic formation of unstable,
nine-membered cores. We have demonstrated YbCl3 to be an
interesting alternative to CeCl3 that can better accommodate the
sterically imposing, macrocyclic framework of 4. It should also be
noted that the final three-step sequence has been reliably
15 A. Bouzide, N. LeBerre and G. Sauve´, Tetrahedron Lett., 2001, 42, 8781.
16 4b was recovered.
17 A simplified model study indicated that Tf2O-induced imidazo[1,5-
a]pyridine formation likely occurred instead of triflate formation, see the
formation of 23 from 22 in supporting information.
18 V. P. Reddy, D. R. Bellew and G. K. S. Prakash, J. Fluorine Chem.,
1992, 46, 165.
19 Y. Koyama, M. J. Lear, F. Yoshimura, I. Ohashi, T. Mashimo and
M. Hirama, Org. Lett., 2005, 7, 267.
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 3057–3059 | 3059