Determination by asymmetric total synthesis of the absolute configuration of lucilactaene, a cell-cycle inhibitor in p53-transfected cancer cells

Article abstract of DOI:10.1002/anie.200500060

(Chemical Equation Presented) A biomimetic pathway to lucilactaene (1) from NG-391 has been developed which involves stereoselective reactions under very mild conditions. It was demonstrated that 1 racemizes rapidly, and the conditions under which racemiz

Full text of DOI:10.1002/anie.200500060

Communications
Natural Products Synthesis
Determination by Asymmetric Total Synthesis of
the Absolute Configuration of Lucilactaene, a
Cell-Cycle Inhibitor in p53-Transfected Cancer
Cells**
Junichiro Yamaguchi, Hideaki Kakeya, Takao Uno,
Mitsuru Shoji, Hiroyuki Osada, and Yujiro Hayashi*
The tumor-suppressor gene p53 is involved in important
cellular events, such as cell-cycle control and apoptosis.[1] The
p53 gene is lost or mutated in many types of human tumors.
Small molecules that induce cell-cycle arrest or apoptosis in a
p53-independent manner or allow mutant p53 to alter a
conformationally active form of p53 may be good candidates
for treating various types of cancers.[2] Recently we isolated
lucilactaene (1), which arrests cell-cycle progression in the
G1 phase at the nonpermissive temperature of 378C in Hl299/
tsp53 cells, from Fusarium sp. RK97-94.[3] Lucilactaene (1) is a
synthetically challenging molecule because of its rare
hexahydro-3a-hydroxy-5-oxo-2H-furo[3,2-b]pyrrol-6-yl ring
system and its substituted and conjugated E,E,E,E,E pen-
taene moiety, which is unstable to acid, base, and light.
Along with lucilactaene (1), we isolated a known neuro-
nal-cell-protecting compound, NG-391 (2),[4] which possesses
the same pentaene portion but a different g-lactam moiety,
and which is probably biosynthesized from the same inter-
mediate as 1.[3] These natural products 1 and 2 are close
structural relatives of fusarins A and C,[5] nonmutagenic
metabolites of Fusarium moniliforme. Though the biosyn-
thetic pathways that lead to 1 and 2 remain unclear, the
following is a plausible pathway based on the proposed
biosynthetic route to fusarin C:[5] The fully elaborated
polyketide reacts with homoserine aldehyde by an intra-
molecular Knoevenagel reaction to form the unmodified 1,5-
dihydropyrrol-2-one 3, a possible common key intermediate
of 1 and 2, after cleavage of the thioester (NADPH reduction)
and condensation (Scheme 1). In the case of NG-391 (2), the
remaining steps are epoxidation and oxidation to form the
hemiaminal, either by ether hydroxylation a to the nitrogen
atom or oxidation to an imine and addition of water (the
[*] J. Yamaguchi, T. Uno, Dr. M. Shoji, Prof. Dr. Y. Hayashi
Department of Industrial Chemistry
Faculty of Engineering, Tokyo University of Science
Kagurazaka, Shinjuku-ku, Tokyo 162-8601 (Japan)
Fax: (+81)3-5261-4631
E-mail: hayashi@ci.kagu.tus.ac.jp
Dr. H. Kakeya, Prof. Dr. H. Osada
Antibiotics Laboratory, Discovery Research Institute
RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198 (Japan)
[**] This work was partially supported by a Grand-in-Aid for Scientific
Research on Priority Areas (A): “Creation of Biologically Functional
Molecules”, from the Ministry of Education, Culture, Sports,
Science, and Technology of Japan. We are indebted to a referee for a
useful suggestion concerning the mechanism of racemization of 1.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
3110
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200500060
Angew. Chem. Int. Ed. 2005, 44, 3110 –3115

Angewandte
Chemie
corresponds to the key biosynthetic intermediate
3. We had already synthesized 2 from 4.[7] The
remaining steps from 4 to lucilactaene (1) would
be hydrolysis of the nitrile group, an intramolec-
ular Michael reaction with the hydroxy group
(R’ = H) as the nucleophile, and functional-group
transformations (Scheme 2). However, all
attempts, including changing of the order in
which the reactions were carried out, were
unsuccessful.
reactions might occur in the reverse order). In the case of
lucilactaene (1), an intramolecular Michael reaction and
oxidation to form the hemiaminal are the remaining reactions
(again the order of reactions might be reversed). Another
possibility for the biosynthesis of 1 is via 2 through intra-
molecular epoxide-ring opening by the primary hydroxy
group, followed by reduction.
Scheme 2. Proposed synthetic approach to lucilactaene from 4.
A similar side chain is also found in epolactaene, a
neuritogenic compound isolated from the fungal strain
Penicillium sp. BM-1689-P.[6] Clarification of the structure–
activity relationships of lucilactaene (1), NG-391 (2), epolac-
taene, and their derivatives is highly desirable for elucidating
their mechanism of action. We completed the first total
syntheses of 2[7] and epolactaene,[8] and developed a biolog-
ically more potent molecule, epolactaene tertiary butyl ester
(ETB).[9] Recently, we revealed that both epolactaene and
ETB bind to human Hsp60 and inhibit Hsp60 chaperon
activity in vitro and in cultured cells,[9] whereas Kobayashi and
co-workers reported that epolactaene is an inhibitor of
mammalian topoisomerases a and b in vitro.[10]
We therefore considered an approach to 1 from NG-391
(2)[7] (Scheme 3). The formation of methyl ether 6, followed
by reductive removal of the epoxide to give an alkene 5, an
The absolute configuration of NG-391 (2) was determined
by us by asymmetric total synthesis.[7] The optical rotation of
lucilactaene (1) is zero in two different solvents (methanol
and chloroform), which indicates the possibility that 1 is
racemic. That 1 should be racemic appears strange when one
considers its structural resemblance with 2. Because of the
interesting biological properties of 1, its lability, and its rare
structure, and because of the puzzle concerning its absolute
configuration, we have investigated its asymmetric total
synthesis by a biomimetic route.
Scheme 3. Retrosynthetic analysis of lucilactaene.
intramolecular Michael reaction, and deprotection would
afford lucilactaene (1). For this approach to be successful the
reactions would have to proceed under mild conditions to
avoid decomposition of the labile pentaene moiety. Methyl
ether formation and the Michael reaction must proceed with
On the basis of the proposed biosynthetic pathway, we
planned to synthesize 1 from the key intermediate 4, which
Scheme 1. Proposed biosynthesis of lucilactaene (1) and NG-391 (2). ACP=acyl carrier protein, NADPH=nicotinamide adenine dinucleotide
phosphate, PCP=peptidyl carrier protein.
Angew. Chem. Int. Ed. 2005, 44, 3110 –3115
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3111

Communications
high diastereoselectivity for 1 to be generated with high
enantiomeric excess.
the facile racemization of 1, the synthesis of optically pure 1
was investigated.
The methyl ether was formed stereoselectively by the
treatment of 2 with a catalytic amount of TsOH·H2O in
MeOH to afford b-methoxide 7 as a single isomer in which
methanol had captured the acyliminium ion intermediate
from the opposite face to that occupied by the epoxide
(Scheme 4).[11] The next planned transformation was the
reductive removal of the epoxide. Although SmI2 is known to
convert a,b-epoxyketones into a,b-unsaturated ketones,[12] in
this case reductive demethoxylation is faster than epoxide
removal. When epoxylactam 7 was treated with SmI2, a 5-(2-
hydroxyethyl)-2-pyrrolidone derivative was formed. After
some experimentation, it was found that the protecting group
on the nitrogen atom of the amide affects the reactivity of the
compound towards reductive demethoxylation. Thus, 7 was
treated with Boc2O in the presence of triethylamine and a
catalytic amount of DMAP to give the bis-Boc-protected
derivative 8 in 69% yield over two steps. The reductive
removal of the epoxide with SmI2 (2 equiv) now proceeded
efficiently at low temperature without affecting the methoxy
group to afford 9 in 75% yield. The two Boc protecting
groups were then removed by treatment with CF3CO2H
(TFA) in CH2Cl2 at room temperature, whereupon a sponta-
neous Michael reaction and the conversion of the methyl
ether into a hydroxy group gave lucilactaene (1) in 28% yield,
along with lucilactaene methyl ether (10) in 55% yield.
Synthetic 1 exhibited identical spectroscopic properties to
those of the natural product (1H NMR, 13C NMR, IR).
The optical rotation of this synthetic lucilactaene (1) was
zero, identical to that of the isolated natural product and
markedly different to that of 10 ([a]D + 36.6 (c = 0.17,
MeOH)). Lucilactaene methyl ether (10) can also be con-
verted into 1 in 60% yield by treatment with TFA in CH2Cl2;
again the optical rotation of the product is zero. The large
difference in optical rotation between 1 and its methyl ether
10 strongly suggests that the lucilactaene (1) formed is
racemic. If racemization occurs during the synthesis, it must
be during the final treatment with acid. To better understand
To avoid possible racemization, the final cleavage of the
hemiaminal protecting group should be conducted under
neutral conditions. To this end we developed a novel
deprotection method. When NG-391 (2) was treated with
PhSeCH2CH2OH[13] in the presence of a catalytic amount of
TsOH·H2O in CH2Cl2 for 3 h at room temperature, phenyl-
selenylethyl ether 11 was formed in 31% yield as a single
isomer; 2 was recovered in 67% yield (Scheme 5). Although
decomposition occurred upon longer treatment of 2 with acid,
the repeated exposure of recovered 2 to acid led to its
conversion into 11 in a high overall yield of 94%. Both the
amide and the hydroxy groups were protected with Boc2O to
afford 12 in 73% yield. The sequence of steps involving
reductive removal of the epoxide with SmI2, removal of the
Boc groups, and the Michael reaction proceeded as efficiently
as for the methyl ether derivative 8 to afford the bicyclic
compound 14 as a single isomer in 36% yield over two steps.
The transformation of the 2-phenylselenylethoxy group into a
hydroxy group could be carried out under mild reaction
conditions in three novel steps: 1) The oxidation of the
selenide to the selenoxide, which was isolated in good yield,
proceeded smoothly at low temperature on treatment with
dimethyldioxirane (DMD),[14] without affecting the pentaene
moiety. 2) The elimination of benzeneselenenic acid occurred
at 608C in the presence of dabco to provide vinyl ether 16.[15]
3) Final oxidative removal of the vinyl substituent was
performed under neutral conditions by the use of DMD at
low temperature (À788C) to afford lucilactaene (1) in 56%
yield from 15 in optically pure form ([a]D + 39.5 (c = 0.10,
MeOH)).[16]
It is clear from the asymmetric total synthesis that isolated
natural lucilactaene (1) is racemic. The facile racemization
raises another question: There is the possibility that natural 1
is optically active but that racemization proceeds during the
purification process. To investigate this possibility, it was
necessary to isolate 1 under nonracemizing conditions. We
Scheme 4. Synthesis of lucilactaene. Boc=tert-butoxycarbonyl, DMAP=4-dimethylaminopyridine, Ts=p-toluenesulfonyl.
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2005, 44, 3110 –3115
3112
www.angewandte.org

Angewandte
Chemie
Scheme 5. Synthesis of optically pure lucilactaene; dabco=1,4-diazabicyclo[2.2.2]octane.
chose to use bicycle 17 as a model compound to establish such
conditions.
Table 1: Racemization of the lucilactaene model 17.
Optically pure 17, prepared by the same method as that
used for 1, was treated with various reagents, and after a
certain period of time the optical purity of the recovered 17
was measured by HPLC analysis on a chiral phase; the results
are summarized in Table 1. Under weakly acidic or basic
conditions, for example, in the presence of PPTS in MeOH or
NEt3 in CH2Cl2, no racemization was observed. However,
racemization occurred when 17 was treated with TFA/CH2Cl2
or K2CO3 in MeOH. It was also confirmed that no racemi-
zation occurred in the medium in which the fermentation was
carried out. These results indicate that the purification of
lucilactaene (1) should be performed under nearly neutral,
mild reaction conditions. The racemization might occur via
intermediates such as 18 or 19, which arise from a reversible
retro-Michael reaction, followed by acyliminium ion forma-
tion or keto–amide formation, though the order of the
reactions might be different (Scheme 6).
As information about the racemization under a variety of
conditions had been obtained, the production profile of
lucilactaene (1) by Fusarium sp. RK97-94 was investigated
further. All experiments were performed as rapidly as
possible, with the temperature and pH value controlled
carefully. The ethyl acetate extracts of the broth (super-
natant) and the mycelia, which were obtained by
centrifugation, were prepared under mild conditions at
pH 7.0. The production profile of 1 in the broth is
summarized in Table 2. The optical purity of 1 in the
broth was very low (ca. 10% ee) throughout the fermen-
tation. Moreover, the lucilactaene (1) in the mycelia was
also nearly racemic (data not shown).
Entry
Reagent
T [8C]
t [h]
ee [%][a]
1
2
3
4
5
6
7
8
9
none
23
23
23
À78
23
0
24
3
3
0.5
3
0.1
0.25
3
3
2.5
100
100
100
100
100
98
57
48
2
0
AcOH/CH2Cl2 (1:20)
PPTS in MeOH (0.005m)
DMD in acetone (0.07m)
Et3N/CH2Cl2 (1:4)
TFA/CH2Cl2 (1:100)
TFA/CH2Cl2 (1:20)
0
TsOH·H2O in CH2Cl2 (0.013m)
K2CO3 in MeOH (0.15m)
TFA/CH2Cl2 (1:4)
23
23
0
10
11
culture medium[b]
28
48
100
[a] Optical purity was determined by HPLC analysis on a chiral phase
(chirapak AD-H). [b] Culture medium: 2% glucose, 1% soluble starch,
0.3% meat extract, 2.5% yeast extract, 0.05% NaCl, 0.005% K2HPO4,
0.05% CaCO3, and 0.05% MgSO4·H2O adjusted to pH 7.2. PPTS=
pyridinium p-toluenesulfonate.
As shown in Scheme 1, lucilactaene (1) and NG-391
(2) may be biosynthesized from the same intermediate 3.
Epoxidation and oxidation to form the hemiaminal
produce 2; these two reactions proceed in this order, as
2 would otherwise be racemic. The absolute configura-
Scheme 6. Racemization of 17 via 18 and/or 19.
Angew. Chem. Int. Ed. 2005, 44, 3110 –3115
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3113

Communications
Table 2: Production profile of 1 by Fusarium sp. RK97-94 in the broth.
affect the labile E,E,E,E,E pentaene moiety; ether formation
from 2 to 7 and 11 and the intramolecular Michael reaction
from 9 to 10 or from 13 to 14 are both highly stereoselective;
the reductive removal of the epoxide with SmI2 without
effecting demethoxylation, and the deprotection of a hemi-
aminal under neutral, oxidative conditions via vinyl ether 16
by using a newly developed phenylselenylethyl protecting
group, are also useful transformations. Detailed biological
studies on both enantiomers of lucilactaene are underway, the
results of which will be reported in due course.
Entry
t [h]
pH[a]
PCV[b] [%]
Production of 1
ee [%][c]
[mgmLÀ1
]
1
2
3
4
24
48
72
96
7.4
7.2
8.3
8.5
10
30
45
50
0
n.d.[d]
10.0
10.3
6.9
0.014
0.084
0.71
[a] pH value of the fermentation broth. [b] Packed-cell volume (v/v).
[c] Optical purity was determined by HPLC analysis on a chiral phase
(chiralcel OD-RH). [d] Not determined.
Received: January 7, 2005
Published online: April 14, 2005
tion of the epoxide can also be explained reasonably as arising
from the selective epoxidation from the opposite face to that
with the hydroxyethyl substituent. In the synthesis of 1, a
Michael reaction and oxidation to form the hemiaminal are
the remaining steps. Should the Michael reaction be the first
step, then we suspect that racemization must occur during the
subsequent oxidation step, because the internal Michael
addition is a spontaneous reaction, as demonstrated for the
synthetic conversion of 13 into 14 without racemization. That
is, there is an acidic moiety in the enzyme responsible for this
oxidation which causes racemization. If oxidation is the first
step, a subsequent racemization process is conceivable
(Scheme 7): Intermediate 3 may undergo tautomerism to a
Keywords: asymmetric synthesis · biosynthesis · lucilactaene ·
.
racemization · total synthesis
[1] a) B. A. Forster, H. A. Coffey, M. J. Morin, F. Rastinejad,
Science 1999, 286, 2507; b) B. Vogelstein, D. Lane, A. J. Levine,
Nature 2000, 408, 307.
[2] a) A. J. Levine, Cell 1997, 88, 323; b) D. Hanahan, R. A.
Weinberg, Cell 2000, 100, 57.
[3] H. Kakeya, S.-I. Kageyama, L. Nie, R. Onose, G. Okada, T.
Beppu, C. J. Norbury, H. Osada, J. Antibiot. 2001, 54, 850.
[4] T. Sugawara, H. Shinonaga, H. Simura, R. Yoshikawa, K.
Yamamoto, Jpn. Kokai Tokkyo Koho 319289, (December 3,
1996).
[5] a) W. C. A. Gelderblom, W. F. O. Marasas, P. S. Steyn, P. G.
Thiel, K. J. Merwe, P. H. Rooyen, R. Vleggaar, P. L. Wessels, J.
Chem. Soc. Chem. Commun. 1984, 122; b) P. S. Steyn, R. J.
Vleggaar, J. Chem. Soc. Chem. Commun. 1985, 1189; c) Z. Song,
R. J. Cox, C. M. Lazarus, T. J. Simpson, ChemBioChem 2004, 5,
1196.
[6] H. Kakeya, I. Takahashi, G. Okada, K. Isono, H. Osada, J.
Antibiot. 1995, 48, 733.
[7] Y. Hayashi, J. Yamaguchi, M. Shoji, Tetrahedron 2002, 58, 9839.
[8] a) Y. Hayashi, K. Narasaka, Chem. Lett. 1998, 313; b) Y.
Hayashi, J. Kanayama, J. Yamaguchi, M. Shoji, J. Org. Chem.
2002, 67, 9443; for total syntheses by other research groups, see:
c) S. Marumoto, H. Kogen, S. Naruto, J. Org. Chem. 1998, 63,
2068; d) S. Marumoto, H. Kogen, S. Naruto, Tetrahedron 1999,
55, 7129; e) S. Marumoto, H. Kogen, S. Naruto, Tetrahedron
1999, 55, 7145; f) K. Kuramochi, H. Itaya, S. Nagata, K.-i.
Takano, S. Kobayashi, Tetrahedron Lett. 1999, 40, 7367; g) K.
Kuramochi, S. Nagata, H. Itaya, K.-i. Takano, S. Kobayashi,
Tetrahedron Lett. 1999, 40, 7371; h) K. Kuramochi, S. Nagata, H.
Itaya, Y. Matsubara, T. Sunoki, H. Uchiro, K.-i. Takao, S.
Kobayashi, Tetrahedron 2003, 59, 9743.
[9] Y. Nagumo, H. Kakeya, J. Yamaguchi, T. Uno, M. Shoji, Y.
Hayashi, H. Osada, Bioorg. Med. Chem. Lett. 2004, 14, 4425.
[10] Y. Mizushima, S. Kobayashi, K. Kuramochi, S. Nagata, F.
Sugawara, K. Sakaguchi, Biochem. Biophys. Res. Commun.
2000, 273, 784; b) J. Nakai, K. Kawada, S. Nagata, K. Kuramochi,
H. Uchiro, S. Kobayashi, M. Ikekita, Biochim. Biophys. Acta
2002, 1581, 1; c) K. Kuramochi, Y. Mizushima, S. Nagata, F.
Sugawara, K. Sakaguchi, S. Kobayashi, Bioorg. Med. Chem.
2004, 12, 1983.
Scheme 7. Possible racemization mechanism in the biosynthesis of
lucilactaene.
2-hydroxypyrrole derivative, which could undergo epoxida-
tion and isomerization. Racemization could occur via the
achiral pyrrole tautomer. A subsequent Michael reaction
would then afford racemic lucilactaene (1).
In summary, the labile natural product lucilactaene (1),
which readily undergoes racemization, has been synthesized
for the first time in optically pure form via a biomimetic
pathway. The conditions under which racemization occurs
were elucidated during this total synthesis. The careful
isolation of lucilactaene (1) from both the broth and the
mycelia under neutral, nonracemizing conditions demon-
strated that the isolable natural product is in fact itself
racemic. This total synthesis, which enabled verification of the
absolute configuration of 1, has several noteworthy features:
All the reactions from NG-391 (2) are mild enough not to
[11] The determination of the stereostructure of 7 is described in the
Supporting Information.
[12] J. M. Concellon, E. Bardales, Org. Lett. 2002, 4, 189.
[13] J. E. Baldwin, R. M. Adlington, M. B. Mitchell, J. Robertson,
Tetrahedron 1991, 47, 5901.
[14] A. Ishii, S. Matsubayashi, T. Takahashi, J. Nakayama, J. Org.
Chem. 1999, 64, 1084.
3114
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 3110 –3115

Angewandte
Chemie
[15] a) L. Engman, J. Org. Chem. 1989, 54, 884; b) K. Haraguchi, H.
Tanaka, H. Maeda, Y. Itoh, S. Saito, T. Miyasaka, J. Org. Chem.
1991, 56, 5401; c) M. Tiecco, L. Testaferri, M. Tingoli, F. Marini,
J. Org. Chem. 1993, 58, 1349, and references therein.
[16] The optical purity was determined by HPLC analysis on a chiral
phase (chiralcel OD-RH column, H2O/CH3CN (100:45),
1.0 mLminÀ1; tR ((À)-1): 27.7 min, tR ((+)-1): 40.2 min).
Angew. Chem. Int. Ed. 2005, 44, 3110 –3115
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3115