Three rings in one step: A quick approach to IKD-8344

Article abstract of DOI:10.1002/anie.201201395

A highly efficient enantioselective total synthesis of the natural antibiotic IKD-8344 is achieved through a convergent route. This route features an otherwise impossible concurrent formation of the THF rings from a linear polyketide precursor through intramolecular O alkylations of mesylates in competition with normally rather facile β elimination and/or α racemization reactions (see scheme, Ms=methanesulfonyl). Copyright

Full text of DOI:10.1002/anie.201201395

Angewandte
Chemie
DOI: 10.1002/anie.201201395
Natural Product Synthesis
Three Rings in One Step: A Quick Approach to IKD-8344**
Yang Zou and Yikang Wu*
IKD-8344 (1, Scheme 1) was first isolated from an unidenti-
fied alkalophilic Actinomycete (strain no. 8344) in the early
1990s by Nishii and co-workers,[1] and later from Streptomyces
to take up an even greater challenge with the concurrent
formation of three THF rings from a highly sensitive linear
precursor. Such an approach would also circumvent the
undesired complications[3] encountered in the assembly of the
preformed single THF ring containing fragments in the
previous total synthesis.
The outline of our retrosynthetic analysis is shown in
Scheme 2. We planned to use Evansꢀ asymmetric aldol
reaction to establish the C2/C3 aldol motif and to form the
Scheme 1. The structure of IKD-8344 with the relative configuration
determined by single crystal X-ray crystallography and the absolute
configuration depicted arbitrarily (Ref. [1b]).
sp. A6792 by Kim and co-workers.[2] This natural dilactone,
which contains a 28-membered ring, represents a worthy
target for synthetic studies, because of its intriguing molecular
architecture and the potent activities against mouse leukemia
(IC50 0.54 ngmLꢀ1) and Trichinella spiralis.[3]
The most striking structural motifs in IKD-8344 are the
2,5-disubstituted THF rings with one of the two substituting
carbon atoms being part of a hidden aldol unit. These
structural motifs are similar to those found in nonactin and
pamamycins. In previous studies[4,5] on the synthesis of similar
THF-containing targets, we developed a novel route to
enantiomerically pure THF rings by using an SN2 type
intramolecular O alkylation at a sterically hindered carbon
atom in b position to a carbonyl group. Such an SN2 reaction
would normally be considered hopeless, because it competes
with the normally very facile a racemization and b elimina-
tion. In the beginning, en route to a synthesis of nonactin, we
realized the formation of one ring at a time. Later, in the
synthesis of pamamycin 621A, we achieved the formation of
two rings in one step. Encouraged by these results, we decided
Scheme 2. The main features of our retrosynthetic analysis. Bn=ben-
zyl, Ms=methanesulfonyl.
THF rings in the presence of the Evans auxiliary. The most
convenient precursor for the seco acid would be the oxazo-
lidinone-masked hydroxy acid (2), which in turn was expected
to form from the highly sensitive/unstable linear precursor 3,
which contains two carbonyl groups, three leaving groups, and
four hydroxy groups, and is thus a compound that may readily
undergo a range of side reactions. The C1 to C21 chain was
planned to be assembled from two smaller fragments, alkenes
4 and 5, through a cross metathesis reaction.
We started the synthesis with the construction of the C3 to
C11 unit (Scheme 3). The asymmetric aldol reaction between
7 (readily derived from 6[6]) and acrolein under the conditions
described by Crimmins et al.[7] afforded the desired enantio-
pure 8 in 86% yield of isolated product. Subsequent cleavage
of the chiral auxiliary with DIBAL-H[7] followed by reaction
with HS(CH2)3SH in the presence of BF3·Et2O gave dithiane
9. The free OH group was then masked as TBS ether before
deprotonation of the dithiane with tBuLi and reaction with
epoxide 11 to give alcohol 12. Use of nBuLi instead of tBuLi
led to lowered yields for 12 (42%) in this case. However, this
disadvantage was partially compensated by the convenience
of performing the reaction at ambient temperature without
the need to add HMPA.
[*] Dr. Y. Zou, Prof. Dr. Y. Wu
State Key Laboratory of Bioorganic and Natural Products
Chemistry, Shanghai Institute of Organic Chemistry
Chinese Academy of Sciences
345 Lingling Road, Shanghai 200032 (China)
E-mail: yikangwu@sioc.ac.cn
[**] This work was supported by the National Basic Research Program of
China (the 973 Program, 2010CB833200) and the National Natural
Science Foundation of China (21172247, 21032002, 20921091,
20621062).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201201395.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Scheme 4. Conversion of 15 into 4. Reagents and conditions: a) 16,
TiCl4, (+)-sparteine, NMP, CH2Cl2, 08C, 90%; b) PIFA, MeOH/H2O,
78%; c) 40% HF, MeCN, 96%. NMP=N-methyl-2-pyrrolidinone,
PIFA=phenyliodine(III) bis(trifluoroacetate).
Scheme 3. Synthesis of 15. Reagents and conditions: a) EtCO2H,
EDCI, DMAP, CH2Cl2, 96%; b) TiCl4, (ꢀ)-sparteine, acrolein, CH2Cl2,
08C, 86%; c) 1) DIBAL-H, CH2Cl2, ꢀ788C; 2) HS(CH2)3SH, BF3·Et2O,
69% from 8; d) TBSOTf, Et3N, CH2Cl2, 08C, 90%; e) tBuLi, THF,
HMPA, ꢀ788C, 71%; f) BnBr, NaHMDS, THF, DMF, 90%; g) CSA,
MeOH, ꢀ108C, 93%; h) SO3-py, DMSO, Et3N, CH2Cl2, 90%. CSA=
camphorsulfonic acid, DIBAL-H=diisobutylaluminum hydride,
DMAP=4-dimethylaminopyridine, DMF=N,N-dimethylformamide,
DMSO=dimethyl sulfoxide, EDCI=1-ethyl-3-(3-dimethylaminopropyl)-
carbodiimide hydrochloride, HMPA=hexamethylphosphoramide,
NaHMDS=sodium hexamethyldisilazide, py=pyridine, TBS=tert-
butyldimethylsilyl, Tf=trifluoromethanesulfonyl.
Benzyl protection of the newly formed hydroxy group at
C6 was unexpectedly difficult. The etherification with BnBr in
the presence of NaH resulted in unwanted side products in
pure DMSO or DMF. Fortunately, this problem was later
satisfactorily solved by using a 5:1 mixture of THF/DMF as
the solvent and performing the alkylation at ꢀ108C.
Furthermore, the incorporation of the C1/C2 unit was
planned by an Evans aldol reaction. Thus, the C3 atom had to
be transformed into an aldehyde first. To this end, the TBS
protecting group at the primary OH group (C3) was
selectively cleaved with MeOH in the presence of a catalytic
amount of CSA, while the TBS group on the secondary OH
group (C10) remained intact. The resulting alcohol 14 was
then treated with SO3·py in CH2Cl2 to deliver the expected
terminal aldehyde 15, thus setting the stage for the incorpo-
ration of the first two carbon atoms in the C1 to C11 fragment
(4).
The aldol reaction was achieved (Scheme 4) under the
conditions described by Crimmins et al.[8] including TiCl4/
sparteine/NMP. The enantiomerically pure 17 was then
treated with PIFA[9] to deprotect the thioketal protecting
group. Finally, removal of the TBS group with HF afforded
alkene 4, which is needed for the later catenation with 5.
The synthesis of fragment 5 was started with the
installation of the C16/C17 stereogenic centers (Scheme 5).
The aldol reaction of 19 with 20 gave 21 in 99% yield. TBS
protection of the hydroxy group with TBSOTf/2,6-lutidine
followed by cleavage of the oxazolidinethione auxiliary with
DIBAL-H gave the intermediate aldehyde.
Scheme 5. Synthesis of 5. Reagents and conditions: a) TiCl4, (+)-spar-
teine, NMP, CH2Cl2, 08C, 99%; b) TBSOTf, 2,6-lutidine, CH2Cl2, 08C,
95%; c) 1) DIBAL-H, CH2Cl2, ꢀ788C, 88%; 2) 23, TiCl4, iPr2NEt,
CH2Cl2, ꢀ788C, 77%; d) 1) MeNH(OMe)·HCl, Me3Al, CH2Cl2;
=
2) (CH2 CH)4Sn, MeLi, THF; 3) Me4NBH(OAc)3, AcOH, MeCN,
ꢀ158C, 71% from 24; e) 1) Me2C(OMe)2, acetone, p-TsOH;
2) nBu4NF, THF, 93% from 25. Ts=toluenesulfonyl.
To extend the chain by two carbon units and install a new
stereogenic center at C15, an asymmetric acetate aldol
reaction that was induced by an oxazolidinethione auxiliary
was performed next. Generally speaking, the diastereoselec-
tivity for such reactions is less satisfactory than that for the
corresponding propionate aldol condensations under similar
conditions. Several elegant protocols[10] for dealing with these
types of problems have been reported. However, the advan-
tages of using these protocols appear to be somewhat
counterbalanced by, for instance, the extra efforts required
for gaining access to the special auxiliaries (with sterically
more hindered substituents at the stereogenic centers). In
hope to find an alternative that is more suitable to us, we
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
decided to examine the aldol condensation of 23, which to our
knowledge has never been examined to date.
desilylation with nBu4NF provided alkene 5, which was
needed for coupling with compound 4.
The results were better than those[11] observed with the
oxazolidinone counterpart of 23 (i.e., with an oxygen atom in
place of the sulfur in 23). Thus, under the same conditions,
including TiCl4/iPr2NEt/CH2Cl2/ꢀ788C, the desired diaste-
reomer 24 was isolated in enantiopure form in 77% yield
(along with 15% of the C15 epimer).
The coupling of 4 with 5 was most satisfactorily achieved
by using 26 as the catalyst with a 1:1.2 molar ratio of 4/5
(Scheme 6). The reaction proceeded rather fast at ambient
temperature (affording 27 in 44% yield within just one hour).
Almost all the excess 5 that was added could be recovered and
recycled. Higher temperature (heating in toluene), which was
normally beneficial, did not lead to any improvements, but
significantly reduced the yield of 27.
Installation of the vinyl group and the stereogenic center
at C13 was mediated by the corresponding Weinreb amide.
The auxiliary in 24 was cleaved by treatment with MeNH-
(OMe)-HCl/Me3Al. However, the subsequent reaction with
vinyl Grignard reagent was complicated by Michael addition
of the MeNH(OMe) onto the vinyl double bond. Exhaustive
optimization efforts included changing the reaction condi-
tions and procedure, but were unsuccessful. Fortunately, the
annoying side reaction could be greatly suppressed under the
=
Saturation of the C C bond in 27 without removing the
Bn groups (which is critical for the selective mesylation) was
much more difficult than the similar reaction in the syn-
thesis[5] of pamamycin 621A. The presence of the carbonyl
group at C8, which greatly facilitated a racemization and
b elimination, led to a failure of the reaction with the protocol
that was previously rather successful (H2/Pd-C/Et3N). Fortu-
nately, after exhaustive tries we found that these problems
could be avoided by using a PtO2-catalyzed hydrogenation.
The resulting triol 28 was converted to the corresponding
trimesylated compound, which on removal of the acetonide
and benzyl groups provided the key precursor 3 (crude) for
the ring-closing reaction. The formation of three THF rings in
conditions described by Fuwa and Sasaki,[12] using (CH2
=
CH)4Sn/MeLi. The desired conjugated enone that was thus
obtained was stereoselectively reduced by using the condi-
[13]
tions described by Evans et al., including Me4NBH(OAc)3
in AcOH/MeCN, thus giving the 1,3-diol 25 in 71% overall
yield (from 24). Further protection with Me2C(OMe)2 and
Scheme 6. Synthesis of 1. Reagents and conditions: a) 26, CH2Cl2, RT, 24 h, 56%; b) H2, PtO2, EtOAc, 89%; c) MsCl, DMAP, CH2Cl2, pyridine,
99%; d) 1) H2, Pd/C, THF; 2) CSA, MeOH; 3) 2,6-lutidine, 1208C, 2 h, 60% from 29; e) TESCl, DMAP, RT, 91%; f) H2O2, LiOH, THF, 08C, 97%;
g) 2, 32, DMAP, Et3N, toluene, RT, 84%; h) 1) TFA, THF; 2) H2O2, LiOH, THF, 08C, 90% from 33; i) 32, DMAP, Et3N, toluene, reflux, 5 h, 68%.
Ms= methanesulfonyl, Mes=2,4,6-trimethylphenyl, TES=triethylsilyl, TFA=trifluoroacetic acid.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[2] E. I. Hwang, B. S. Yun, W. H. Yeo, S. H. Lee, J. S. Moon, Y. K.
Kim, S. J. Lim, S. U. Kim, J. Microbiol. Biotechnol. 2005, 15,
909 – 912 (No optical rotation data was given).
one step was then achieved with 2,6-lutidine under heating.
Gratifyingly, 2 was obtained in 60% overall yield (from 29).
Coupling of 2 with 31 (prepared from 2 by protection with
a TES group and removal of the chiral auxiliary) under the
Yamaguchi conditions provided 33, which on removal of the
TES group and the chiral auxiliary gave 34. Lactonization of
[3] For the total synthesis, see: a) W. H. Kim, S. K. Hong, S. M. Lim,
M.-A. Ju, S. K. Jung, Y. W. Kim, J. H. Jung, M. S. Kwon, E. Lee,
Angew. Chem. 2006, 118, 7230 – 7233; Angew. Chem. Int. Ed.
2006, 45, 7072 – 7075; b) W. H. Kim, S. K. Hong, S. M. Lim, M.-
A. Ju, S. K. Jung, Y. W. Kim, J. H. Jung, M. S. Kwon, E. Lee,
Tetrahedron 2007, 63, 9784 – 9801. For the synthesis of fragments,
see: c) W. Jiang, D. A. Lantrip, P. L. Fuchs, Org. Lett. 2000, 2,
2181 – 2184.
[4] Y.-K. Wu, Y.-P. Sun, Org. Lett. 2006, 8, 2831 – 2834.
[5] a) G.-B. Ren, Y.-K. Wu, Org. Lett. 2009, 11, 5638 – 5641; b) G.-B.
Ren, Y.-X. Huang, Y.-P. Sun, Z.-H. Li, Y.-K. Wu, J. Org. Chem.
2010, 75, 5048 – 5064; c) G.-B. Ren, Y.-K. Wu, Chin. J. Chem.
2010, 28, 1651 – 1659.
[6] For a practical access to 6, see: Y.-K. Wu, Y.-Q. Yang, Q. Hu,
J. Org. Chem. 2004, 69, 3990 – 3992.
[7] a) M. T. Crimmins, K. Chaudhary, Org. Lett. 2000, 2, 775 – 777;
b) M. T. Crimmins, B. W. King, E. A. Tabet, K. Chaudhary,
J. Org. Chem. 2001, 66, 894 – 902.
[8] M. T. Crimmins, J. She, Synlett 2004, 1371 – 1373.
[9] G. Stork, K. Zhao, Tetrahedron Lett. 1989, 30, 287 – 290.
[10] a) N. R. Guz, A. J. Phillips, Org. Lett. 2002, 4, 2253 – 2256; b) Y.
Zhang, T. Sammakia, Org. Lett. 2004, 6, 3139 – 3141; c) M. T.
Crimmins, M. Shamszad, Org. Lett. 2007, 9, 149 – 152.
[11] C. Le Sann, D. M. Munoz, N. Saunders, T. J. Simpson, D. I.
Smith, F. Soulas, P. Watts, C. L. Willis, Org. Biomol. Chem. 2005,
3, 1719 – 1728.
1
34 afforded product 1, the H and 13C NMR spectra and the
27
specific rotation (½aꢁD ¼ + 36.33 (c = 0.20, MeOH), lit.[14]
17
½aꢁD ¼ + 40.2 (c = 1.0, MeOH)) of which were consistent
with those of the natural product 1.
Only the relative configuration of natural product 1 was
ever experimentally determined, although the absolute con-
figuration that was arbitrarily depicted in the report of the
structure[1b] happened to be correct. In retrospect, the first
synthesis[3a,b] of 1 by Lee and co-workers would have allowed
to establish the absolute configuration. But unfortunately, the
Lee group measured the rotation of 1 in CHCl3 rather than
MeOH,[15] the absolute configuration of natural product
1 thus remained unelucidated to date.
In summary, an efficient synthesis of IKD-8344 has been
achieved with a novel “three rings in one step” transformation
as the key step. Apart from the formation of multiple THF
rings in an unprecedented difficult and complex situation, the
=
efficient chain assembly and the tricky saturation of the C C
bond in a highly sensitive substrate are also interesting. The
absolute configuration of natural product 1, which was
arbitrarily depicted in the original report of the configura-
tion[1b] and mistaken as proven without being noticed to date,
is also verified for the first time; the corresponding hidden
and potential confusions in the literature and the major online
data bases are thus cleared.
[12] H. Fuwa, M. Sasaki, Org. Lett. 2010, 12, 584 – 587.
[13] D. A. Evans, K. T. Chapman, E. M. Carreira, J. Am. Chem. Soc.
1988, 110, 3560 – 3578.
[14] A full description of the specific rotation for the natural IKD-
8344 can be found in: M. Nishii (Ikeda Mohando Co., Ltd.) Jpn.
Kokai Tokkyo Koho, JP 02009382, 1990; Chem. Abstr. 1990, 113,
103380. This patent appeared to have been overlooked during
the previous synthetic studies because only the incomplete
rotation data (without the solvent) from Ref. [1a] was ever cited.
[15] Comparison of rotations measured in different solvents cannot
provide any reliable clues to the absolute configuration. For
instance, pamamycin 621A, also a THF-containing natural
product, gave opposite rotations in CHCl3 and CH2Cl2
(Ref. [5a,b]). The rotation for 1 in CHCl3 from this work is
Received: February 20, 2012
Published online: && &&, &&&&
Keywords: cyclizations · eliminations · natural products ·
.
ring-closing reactions · total synthesis
27
½aꢁD ¼ + 20.2 (c = 0.38, CHCl3), and thus remarkably different
15
from that in Ref. [3a,b] (½aꢁD ¼ + 39.7 (c = 0.25, CHCl3), the only
[1] a) Y. Minami, K. Yoshida, R. Azuma, M. Nishii, J. Inagaki, F.
Nohara, Tetrahedron Lett. 1992, 33, 7373 – 7376; b) T. Ishida, Y.
In, M. Nishii, Y. Minami, Chem. Lett. 1994, 1321 – 1322.
complete rotation data for 1 (erroneously registered with a never
proven absolute configuration) that can be found in leading
online databases such as SciFinder and Reaxys to date).
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Natural Product Synthesis
Y. Zou, Y. Wu*
&&&& — &&&&
Three Rings in One Step: A Quick
Approach to IKD-8344
A highly efficient enantioselective total
synthesis of the natural antibiotic IKD-
8344 is achieved through a convergent
route. This route features an otherwise
impossible concurrent formation of the
THF rings from a linear polyketide pre-
cursor through intramolecular O alkyla-
tions of mesylates in competition with
normally rather facile b elimination and/
or a racemization reactions (see scheme,
Ms=methanesulfonyl).
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!