Expeditious synthesis of hippuristanol and congeners with potent antiproliferative activities

Article abstract of DOI:10.1002/chem.200901732

Rapid access : An expeditious synthesis of hippuristanol was developed that allowed rapid access to a number of analogues with structural alteration at its E and F rings (see scheme), facilitating the structure-activity relationship studies of the novel i

Full text of DOI:10.1002/chem.200901732

DOI: 10.1002/chem.200901732
Expeditious Synthesis of Hippuristanol and Congeners with Potent
Antiproliferative Activities
Wei Li,[a] Yongjun Dang,[b] Jun O. Liu,*[b] and Biao Yu*[a]
Hippuristanol (1) and its congeners comprise a family of
lithio-2,2,3-trimethyl-2,3-dihydrofuran (A) to a 5a-pregn-
structurally unique polyoxygenated steroids isolated from
the Gorgonian Isis hippuris with potent antiproliferative ac-
tivity against different cancer cell lines in vitro.[1] The molec-
ular basis of their antiproliferative activity, however, re-
mained a mystery until recently. Hippuristanol was reported
to target the eukaryotic translation initiation factor
(eIF)4A,[2a,b] an ATP-dependent RNA helicase that plays a
pivotal role in translation in eukaryotic cells. The discovery
of hippuristanol, along with another marine sponge-derived
natural product pateamine A, as inhibitors of eukaryotic
translation, not only offered new molecular probes to study
eukaryotic translation initiation mediated by eIF4A, but
also had important implications in targeting the translation
process for the development of novel anticancer and antivi-
ral drugs.[2] Fascinated by its unusual chemical architecture,
especially the bicyclic spiroketal appendage (rings E and F)
to the conventional steroid core, and its novel biological ac-
tivity, we embarked on the synthesis of hippuristol. Herein,
we report the first synthesis of hippuristanol and its conge-
ners with structural and stereochemical variations on E and
F rings, along with their characterization in both cell prolif-
eration and in vitro translation assays.
3a,11b,16b-trihydroxy-20-one derivative, such as
B
(Scheme 1);[3] the resulting adduct would have all skeletal
Scheme 1. Hippuristanol (1) and 22-epi-hippuristanol (2), and the retro-
synthetic analysis.
carbons and functional groups in place for further elabora-
tion. The nascent stereochemistry (at C20 and C22) would
be determined largely by the pre-existing stereocenters on
the steroidal substrate B. Thus, modification of the sub-
strates to effect the natural stereochemistry might be re-
quired. The control, or lack of it, of the stereochemical out-
come of the reaction leading to different stereoisomers
could also be taken advantage of to explore the structure–
activity relationship of the E and F rings of hippuristanol. In
addition, by employing a variety of the 2,3-dihydrofuran de-
rivatives (such as 2,3-dihydrofuran 18 and 2,2-dimethyl-2,3-
dihydrofuran 19), we could access the corresponding conge-
ners with altered substitutions at the F ring. Hydrocortisone
(3) was chosen as the starting precursor to the steroidal sub-
strate B due to its commercial availability and low cost. It
possesses the steroidal ABCD-ring skeleton and the re-
quired oxy-function (of B) at C3, C20, and C11. In particu-
lar, the b-OH at C11 of 3 would otherwise be difficult to in-
troduce onto steroids core structure by conventional chemis-
try.
In a retrosynthetic analysis, we envisioned an expeditious
approach to hippuristanol (1) via addition of racemic 5-
[a] W. Li, Prof. Dr. B. Yu
State Key Laboratory of Bio-organic and
Natural Products Chemistry
Shanghai Institute of Organic Chemistry
Chinese Academy of Sciences, 354 Fenglin Road
200032 Shanghai (China)
Fax : (+86)21-64166128
E-mail: byu@mail.sioc.ac.cn
[b] Dr. Y. Dang, Prof. Dr. J. O. Liu
Department of Pharmacology and Department of Oncology
Johns Hopkins School of Medicine, 725 North Wolfe Street
Baltimore, 21205 MD (USA)
E-mail: joliu@jhu.edu
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901732.
10356
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10356 – 10359

COMMUNICATION
The synthesis commenced
with conversion of hydrocorti-
sone (3) into the 3a,11b,16b-
trihydroxy-5a-pregn-20-one
derivative
13
(Scheme 2).
Treatment of hydrocortisone
(3) with lithium in liquid NH3
led to a complex mixture.[4]
This could be attributed to un-
desirable reactions on the hy-
droxyacetyl-carbinol residue at
C17, which was thus protected
with
a robust bismethylene-
dioxy function to give com-
pound 4.[5] Reduction of 4 with
lithium in liquid NH3 provided
the desired 5a-H-3-one deriva-
tive 5 in good yield. Ketone 5
was then subjected to reduc-
tion with K-selectride to pro-
vide 3a-ol 6 stereoselectively
(80%).[6,7] Removal of the bis-
Scheme 2. Synthesis of the 3a,11b,16b-trihydroxy-5a-pregn-20-one (13).
methylenedioxy
protecting
group in 6 with formic acid re-
sulted in the elimination of the 11b-OH to give the corre-
sponding D9,11 derivatives.[5] Therefore, the 11b-OH (in 6)
was subjected to protection before further elaboration.
Treatment of 6 with benzoyl chloride under harsh conditions
(reflux in pyridine in the presence of DMAP) provided the
3,11-di-O-benzoyl derivative 7 (99%); under milder condi-
tions, the 3-OH (in 6) could be selectively benzoylated. Re-
moval of the bismethylenedioxy protection from 7 with
formic acid (80% in H2O at 808C) afforded compound 8 to-
gether with its 21-formate derivative, which was converted
to 8 with K2CO3 in a mixed solvent of THF and MeOH at
low temperature (08C). Removal of the 21-OH in 8 was
achieved via tosylate formation, giving 9 (91%), and re-
placement of the tosylate with iodide, followed by reduction
(NaI, acetone, reflux; then AcOH, 658C),[8] furnishing the
desired 10 in 98% yield. Under the action of semicarbazide
in aqueous acetic acid at 808C,[8c,9] the a-hydroxy-ketone 10
was converted to enone 11 in 83% yield. A rearranged com-
pound (S1)[10] was detected in trace amount, which had been
the major product upon treat-
with difficulties. Michael addition or hydroboration–oxida-
tion of enone 11 under a variety of conditions provided the
16-a-hydroxy products.[12c,13] Fortunately, under the action of
N-bromoacetamide (NBA),[12] enone 11 could be trans-
formed into the 16b-hydroxyl-17a-bromide 12 in 69% yield.
Removal of the 17-bromide under radical conditions
(Bu3SnH, Et3B, CH2Cl2, RT) afforded the desired 16b-hy-
droxy-20-ketone 13 cleanly (98%).
The volatile 2,2,3-trimethyl-2,3-dihydrofuran (16) was
readily prepared from the commercially available a-methyl-
g-butyrolactone 14 in four steps and in 30% overall yield
(Scheme 3).[3c,14] Treatment of ketone 13 with dihydrofuran
16 in the presence of tBuLi in THF at 08C followed with
0.1m HCl at RT provided a rather complex mixture, partly
because of partial cleavage of the two O-benzoyl groups.
Nevertheless, complete removal of the benzoyl groups with
LiAlH4 (THF, 408C) afforded three stereoisomeric products,
which were characterized as 22,24-di-epi-hippuristanol (17a)
(18%), 24-epi-hippuristanol (17b) (22%), and 22-epi-hip-
ment of 10 with SOCl2 in pyri-
dine at room temperature.[11] It
is worth noting that a substrate
without the benzoyl protection
at the 11-OH (namely, 3a-ben-
zoyloxy-11b,17a-dihydroxy-5a-
pregn-20-one) gave rise to the
corresponding enone and rear-
ranged product in nearly 1: 1
ratio upon treatment with sem-
icarbazide (data not shown).
Introduction of the 16-b-OH
onto enone 11 was initially met
Scheme 3. Elaboration of hippuristanol (1), 22-epi-hippuristanol (2), and stereoisomers 17a/17b.
Chem. Eur. J. 2009, 15, 10356 – 10359
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10357

J. O. Liu, B. Yu et al.
puristanol (2) (43%), in 83% overall yield. Hippuristanol
(1) was obtained in ꢀ40% yield via epimerization of 22-epi-
hippuristanol 2 under the action of a catalytic amount of
TsOH in CH2Cl2 at RT. The corresponding D9,11 product
from elimination of the 11-OH was not detected (cf. the
acidic treatment of compound 6).
The same synthetic route to hippuristanol was easily
adapted to the synthesis of a variety of hippuristanol conge-
ners (Scheme 4). Thus, addition of 22-ketone 13 with 2,3-di-
hydrofuran 18 (tBuLi, THF, À688C) followed by further
elaboration (0.1m HCl, RT; then LiAlH4, THF, 408C) pro-
vided a pair of the C22 epimers 20a and 20b in 38 and 28%
yield, respectively. Compounds 20a and 20b lack the three
methyl groups of hippuristanols on the F ring. Similarly,
treatment of 13 with 2,2-dimethyl-2,3-dihydrofuran (19)
through three steps afforded 21a (26%) and 21b (51%),
which lack the 24-methyl group of the natural product.
408C) furnished the final products (22a/b, 23, and 24a/b) in
reasonable yields. The C22(S)-isomer 22b resulted from the
C22(R) 22a upon acidic workup (5% HCl, RT).
Assignment of the structures of these synthetic congeners
(1, 2, 17a/b, 20a/b, 21a/b, 22a/b, 23, and 24a/b) proved to
be extremely difficult by spectroscopic methods. Fortunately,
the natural product hippuristanol (1) and 22-epi-hippurista-
nol (2) have been well characterized by extensive spectro-
scopic analysis as well as correlation with an X-ray structure
1
of a natural congener (hippurin-1).[1] The H and 13C NMR
spectra of the synthetic compounds 1 and 2 are identical to
those of the authentic natural products. Besides, the struc-
tures of 22a/b and 23 have been determined by X-ray dif-
fraction of the relevant derivatives (compounds S3 and
S5).[7] Based on these unambiguous assignments, the struc-
tures of the synthetic congeners could be well correlated by
NMR signals of the H16 and C22, as well as the [a]D values
(Table S1).[10]
With synthetic hippuristanol
and congeners in hand, we de-
termined their effects on the
proliferation of transformed
cancer cell line HeLa and on
protein synthesis in vitro. In
general, the two types of activ-
ities are in agreement with
Scheme 4. Divergent assembly of hippuristanol analogues 20a/b and 21a/b. a) tBuLi, THF, À688C (or 08C for
13 + 19). b) 0.1m HCl, RT. c) LiAlH4, THF, 408C.
each other with synthetic hip-
puristanol exhibiting the high-
est activity in both assays. The
Contrary to the addition of 5-lithio-2,3-dihydrofurans to
16b-OH-22-one 13, where the 20b-OH adducts predominat-
ed, their addition to enone 11 led to the corresponding 20a-
OH products stereoselectively under similar conditions (a/b
8:1, 19:1, and 1:0, respectively, for addition with 18, 19, and
16) (Scheme 5).[10] The resulting D16,17-22-one-25-ols under-
went cyclization under the action of NIS,[15] providing the
corresponding 16b-O, 17a-iodide, 22-epi-derivative as the
sole stereoisomer, respectively. Subsequent removal of the
17-iodide (Bu3SnH, AIBN, toluene, 708C) followed by de-
protection of the 3,11-O-benzoyl groups (LiAlH4, THF,
synthetic hippuristanol inhibited HeLa cell proliferation
with an IC50 of 72 nm, significantly lower than that (ca.
800 nm) reported previously for the natural product.[2a]
A
clear structure–activity relationship also emerged. First of
all, an “R” configuration at C22 appears to be essential for
the activity. Thus, inversion of the stereochemistry at C22 in
hippuristanol (2) led to a significant decrease in activity
(Table 1). A similar decrease in or loss of activity was also
seen between congeners 21a and 21b, and between 17a and
17b. Given the opposite configuration at the D/E ring junc-
ture and the relative rigidity of the E/F spiroketal functional
group, it is tempting to specu-
late that the “R” stereochemis-
try at C22 would place the
three oxygen atoms on the
same side of a concave that are
capable of multiple hydrogen-
bonding interactions. Second,
of the three methyl groups on
the F ring, which occupy the
convex surface opposite to the
oxygen-rich concave, at least
two are required for activity.
Thus, congeners 20a, 20b, 22a
and 22b are all inactive. In
contrast, 21b that contains the
gem-dimethyl substitution on
the F ring, is quite potent, sug-
Scheme 5. Divergent assembly of 20-epi-hippuristanol analogues 22a/b, 23, and 24a/b. a) tBuLi, THF, À688C.
b) 0.1m HCl, RT. c) LiAlH4, THF, 408C. d) NIS, CH2Cl2, RT. e) Bu3SnH, AIBN, toluene, 708C. f) LiAlH4,
THF, 408C; then 5% HCl, RT.
10358
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 10356 – 10359

Expeditious Synthesis of Hippuristanol
COMMUNICATION
Table 1. Effects of synthetic hippuristanol and its analogues on HeLa
cell proliferation and in vitro translation.
2008, 3, e1583; c) W. K. Low, Y. Dang, T. Schneider-Poetsch, Z. Shi,
N. S. Choi, W. C. Merrick, D. Romo, J. O. Liu, Mol. Cell 2005, 20,
709–722; d) M. E. Bordeleau, J. Matthews, J. M. Wojnar, L. Lindqv-
ist, O. Novac, E. Jankowsky, N. Sonenberg, P. Northcote, P. Tees-
dale-Spittle, J. Pelletier, Proc. Natl. Acad. Sci. USA 2005, 102,
10460–10465; e) J. Clardy, ACS Chem. Biol. 2006, 1, 17–19; f) T. V.
Pestova, C. U. Hellen, Nat. Chem. Biol. 2006, 2, 176–177.
[3] a) Z. Szendi, F. Sweet, Steroids 1991, 56, 458–463; b) U. Hedtmann,
R. Klintz, K. Hobert, J. Frelek, I. Vlahov, P. Welzel, Tetrahedron
1991, 47, 3753–3772; c) U. Hobert, K. Hedtmann, R. Klintz, P.
Welzel, J. Frelek, M. Strangmann-Diekmann, A. Klçne, O. Pongs,
Angew. Chem. 1989, 101, 1543–1546; Angew. Chem. Int. Ed. Engl.
1989, 28, 1515–1518.
Compound
IC50 [mm] HeLa
IC50 [mm] in vitro translation
1
2
17a
0.072
3.59
1.52
0.13
>20
>20
0.36
>20
>20
>20
>20
>20
0.20
14.45
3.24
0.96
N/A
N/A
0.60
N/A
N/A
>50
>50
>50
17b
20a/20b 1:1
21a
21b
22a
22b
23
24a
[4] a) A. L. Nussbaum, T. L. Popper, E. P. Oliveto, S. Friedman, I.
Wender, J. Am. Chem. Soc. 1959, 81, 1228–1231; b) P. F. Beal, M. A.
Rebenstorf, J. E. Pike, J. Am. Chem. Soc. 1959, 81, 1231–1234; c) Y.
Shu, S. R. Jones, W. A. Kinney, B. S. Selinsky, Steroids 2002, 67,
291–304.
24b
[5] a) J. A. Edwards, M. C. Calzada, A. Bowers, J. Med. Chem. 1964, 7,
528–530; b) R. E. Beyler, F. Hoffman, R. M. Moriarty, L. H. Sarett,
J. Org. Chem. 1961, 26, 2421–2425.
[6] a) H. Hosoda, W. Takasaki, H. Miura, M. Tohkin, Y. Maruyama, T.
Nambara, Chem. Pharm. Bull. 1985, 33, 4281–4287; b) A. R.
McKinney, D. D. Ridley, P. Turner, Aust. J. Chem. 2003, 56, 829–
838.
[7] The stereochemistry was confirmed by an X-ray diffraction analysis,
see the Supporting Information.
[8] a) J. Fried, K. Florey, E. F. Sabo, J. E. Herz, A. R. Restivo, A.
Borman, F. M. Singer, J. Am. Chem. Soc. 1955, 77, 4181–4182; b) R.
Vitali, R. Gardi, A. Ercoli, Gazz. Chim. Ital. 1966, 96, 1115; c) R. E.
Conrow, J. Org. Chem. 1999, 64, 1042–1044.
[9] D. Taub, R. D. Hoffsommer, N. L. Wendler, J. Org. Chem. 1964, 29,
3486–3495.
[10] See the Supporting Information for details.
[11] a) S. Bernstein, J. J. Brown, L. I. Feldman, N. E. Rigler, J. Am.
Chem. Soc. 1959, 81, 4956–4962; b) W. S. Allen, S. Bernstein, J. Am.
Chem. Soc. 1955, 77, 1028–1032.
[12] a) B. Loken, S. Kaufmann, G. Rosenkranz, F. Sondheimer, J. Am.
Chem. Soc. 1956, 78, 1738–1741; b) S. G. Levine, M. E. Wall, J. Am.
Chem. Soc. 1959, 81, 2829–2830; c) J. C. Danilewicz, W. Klyne, J.
Chem. Soc. 1965, 1306–1319.
gesting that those methyl groups are likely involved in hy-
drophobic interactions with the target eIF4A.
In summary, an expeditious route was established to syn-
thesize the natural product hippuristanol and its congeners
with structural alterations in its unique E and F rings. The
determination of the biological activity of the synthetic ana-
logues revealed the importance of both stereochemistry of
the spiroketal group and the methyl substitutions in the F
ring. These results allow future synthesis of new analogues
to improve the potency of this family of inhibitors of eu-
karyotic translation initiation and molecule probes to fur-
ther delineate the molecular recognition of hippuristanol by
eIF4A.
Acknowledgements
We thank Prof. Carlos Jimꢁnez of the Universidade de A CoruÇa (Spain)
for providing NMR spectra of the authentic hippuristanol and 22-epi-hip-
puristanol. Financial support from the National Natural Science Founda-
tion of China (20728202 and 20621062), the Chinese Academy of Scien-
ces (KJCX2-YW-H08), and the Committee of Science and Technology of
Shanghai (07DZ22001) are gratefully acknowledged.
[13] a) H. Hirschmann, F. B. Hirschmann, J. W. Corcoran, J. Org. Chem.
1955, 20, 572–586; b) D. K. Fukushima, T. F. Gallagher, J. Am.
Chem. Soc. 1951, 73, 196–201; c) J. Pataki, G. Rosenkranz, C. Djer-
assi, J. Am. Chem. Soc. 1952, 74, 5615–5616; d) M. Nussim, Y.
Mazur, F. Sondheimer, J. Org. Chem. 1964, 29, 1120–1131; e) P.
Tang, B. Yu, Angew. Chem. 2007, 119, 2579–2582; Angew. Chem.
Int. Ed. 2007, 46, 2527–2530.
Keywords: cancer · natural products · steroids · structure–
[14] a) N. C. Barua, R. R. Schmidt, Synthesis 1986, 891–893; b) C. Botte-
ghi, G. Consiglio, G. Ceccarelli, A. Stefani, J. Org. Chem. 1972, 37,
1835–1837; c) L. A. Paquette, J. C. Lanter, J. N. Johnston, J. Org.
Chem. 1997, 62, 1702–1712; d) G. M. Coppola, H. F. Schuster, J.
Heterocycl. Chem. 1984, 21, 1409–1410.
[15] a) S. Kaufmann, L. Tokes, J. W. Murphy, P. Crabbe, J. Org. Chem.
1969, 34, 1618–1621; b) C. A. Townsend, Y. Isomura, S. G. Davis,
J. A. Hodge, Tetrahedron 1989, 45, 2263–2276; c) R. L. Rogers, J. L.
Moore, T. Rovis, Angew. Chem. 2007, 119, 9461–9464; Angew.
Chem. Int. Ed. 2007, 46, 9301–9304.
activity relationships
[1] a) T. Higa, J.-i. Tanaka, Y. Tsukitani, H. Kikuchi, Chem. Lett. 1981,
1647–1650; b) C. B. Rao, K. V. Ramana, D. V. Rao, E. Fahy, D. J.
Faulkner, J. Nat. Prod. 1988, 51, 954–958; c) N. Gonzꢂlez, M. A.
Barral, J. Rodriguez, C. Jimenez, Tetrahedron 2001, 57, 3487–3497.
[2] a) M. E. Bordeleau, A. Mori, M. Oberer, L. Lindqvist, L. S. Chard,
T. Higa, G. J. Wagner, G. Belsham, J. Tanaka, J. Pelletier, Nat.
Chem. Biol. 2006, 2, 213–220; b) L. Lindqvist, M. Oberer, M. Rei-
barkh, R. Cencic, M. E. Bordeleau, E. Voqt, A. Marintchev, J.
Tanaka, F. Fagotto, M. Altmann, G. Wagner, J. Pelletier, PLoS One
Received: June 23, 2009
Published online: September 11, 2009
Chem. Eur. J. 2009, 15, 10356 – 10359
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10359