Enantioselective total synthesis of the selective PI3 kinase inhibitor liphagal

Article abstract of DOI:10.1021/ol101173w

The enantioselective total synthesis of liphagal, a selective inhibitor of PI3K α isolated from the marine sponge Aka coralliphaga, has been achieved. The novel tetracyclic liphagane skeleton is formed in one step, after the hydrogenation of a dihydroxydrimane phenol benzyl ether in the presence of cationic resin.

Full text of DOI:10.1021/ol101173w

ORGANIC
LETTERS
2010
Vol. 12, No. 20
4450-4453
Enantioselective Total Synthesis of the
Selective PI3 Kinase Inhibitor Liphagal
Enrique Alvarez-Manzaneda,* Rachid Chahboun, Esteban Alvarez, Ma Jose´ Cano,
Ali Haidour, and Ramo´n Alvarez-Manzaneda†
Departamento de Qu´ımica Orga´nica, Facultad de Ciencias, Instituto de Biotecnolog´ıa,
UniVersidad de Granada, 18071 Granada, Spain, and Departamento de Qu´ımica
Orga´nica, Facultad de Ciencias, UniVersidad de Almer´ıa, 04120 Almer´ıa, Spain
eamr@ugr.es
Received May 21, 2010
ABSTRACT
The enantioselective total synthesis of liphagal, a selective inhibitor of PI3K r isolated from the marine sponge Aka coralliphaga, has been
achieved. The novel tetracyclic “liphagane” skeleton is formed in one step, after the hydrogenation of a dihydroxydrimane phenol benzyl ether
in the presence of cationic resin.
In recent decades, marine organisms appear to have become
an almost inexhaustible source of natural products, showing
very different structural patterns and a wide variety of
interesting biological activities.1 A paradigmatic example of
this type of metabolites is liphagal (1, Figure 1), a merot-
erpenoid recently isolated from the marine sponge Aka
coralliphaga,2 which exhibits the novel “liphagane” carbon
skeleton. Besides its uncommon structure, liphagal (1)
presents considerable therapeutic potential, with inhibitory
activity against PI3K R (phosphoinositide-3-kinase R). It is
more potent than the synthetic LY 294002 and more selective
than wortmanin, making it a promising candidate as an agent
for the treatment of inflammatory and autoimmune disorders
as well as cancer and cardiovascular diseases.3 In fact,
Figure 1. Metabolites from Aka coralliphaga.
liphagal (1) has been observed to be cytotoxic, in secondary
in vitro assays, to LoVo (human colon: IC50 0.58 µM), CaCo
† Universidad de Almer´ıa.
(human colon: IC50 0.67 µM), and MDA-468 (human breast:
IC50 1.58 µM) tumor cell lines. The protein kinase C
inhibitors corallidictyals, such as corallidictyal B (2) and D
(3), are spirosesquiterpene aldehydes closely related to
compound 1 and isolated from the same natural source.4
(1) For a recent review on marine natural products, see: Blunt, J. W.;
Copp, B. R.; Munro, M. H. G.; Northcote, P. T.; Prinsep, M. R. Nat. Prod.
Rep. 2010, 27, 165.
(2) Marion, F.; Williams, D. E.; Patrick, B. O.; Hollander, I.; Mallon,
R.; Kim, S. C.; Roll, D. M.; Feldberg, L.; Soest, R. W.; Andersen, R. J.
Org. Lett. 2006, 8, 321.
10.1021/ol101173w  2010 American Chemical Society
Published on Web 06/02/2010

Two racemic syntheses of liphagal (1) have been reported.
Andersen et al., after isolating compound 1 from its natural
source, developed a synthesis of this bioactive metabolite,
which involves as the key step the cation-initiated cyclization
of a dienyl benzofuran.2 Recently, Mehta et al. described a
closely related strategy to synthesize (()-1, based on the
acid-promoted cyclization of a cyclohexenyl benzofuran.5
Andersen’s group proposed two possible biogenic path-
ways to liphagal (1), starting from a farnesyl trihydroxy-
benzaldehyde. One of these pathways takes place via the
bicyclofarnesyl trihydroxybenzaldehyde siphonodictyal B
(4),6 a metabolite also found in the sponge Aka coralliphaga.
In the alternative pathway, the benzofuran system is first
formed from an acyclic ketone and the 6,7-ring system is
subsequently created, after the acid promoted cyclization of
a dienyl benzofuran.2
Scheme 2. Synthesis of Epoxy Alcohols 7a,b
However, the stereoselective introduction of the epoxide
function involves certain difficulties.
Scheme 1. Retrosynthesis of Liphagal (1)
It is well-known that the epoxidation of the allyl alcohol
precursor of aldehyde 11 is not stereoselective, leading to a
2.5:1 mixture of the corresponding R- and ꢀ-epoxy deriva-
tives, respectively.10 In order to improve the efficiency of
synthetic sequence, the utilization of this allyl alcohol as an
intermediate was ruled out, and the diene 9 was investigated
as an alternative precursor; this compound underwent chemo-
and stereoselective epoxidation at low temperatures, affording
the epoxy alkene 10 in high yield. Diene 9 has been easily
synthesized from various starting materials, such as a (S)-
(+)-Wieland-Miescher ketone analogue11 or the natural
monoterpene (R)-(-)-carvone;12 compound 9 has also been
obtained after dehydrohalogenation of allyl iodide 8, syn-
thesized after the lipase-catalyzed kinetic resolution following
the acid cyclization of homofarnesyl acetate13 or from
commercial sclareolide, the latter being the most efficient
procedure for synthesizing diene 9 (three steps, 65% overall
yield).14 Ozonolysis of epoxy alkene 10, whose relative
stereochemistry was established on the basis of NOE
experiments, gave aldehyde 11 in good yield.
During our research into the synthesis of bioactive natural
products, we were interested in developing an enantioselec-
tive synthesis of liphagal (1), making it possible to establish
its absolute stereochemistry. Scheme 1 shows the retrosyn-
thesis planned for compound 1. The furan ring will be formed
by dehydration of the hemiketal resulting from the hydroxy
ketone derived from compound 5. This will result from the
pinacol rearrangement of diol 6, resulting from the regiose-
lective reduction of epoxy alcohol 7. This will be obtained
after the addition of an aryllithium to an epoxy aldehyde.
Scheme 2 shows the synthesis of epoxy alcohols 7a,b via
epoxy aldehyde 11. The synthesis of compound 11 is not a
trivial task. Different synthetic procedures have been reported
fortheconstructionofthe(2,5,5,8a-tetramethyl-3,4,4a,5,6,7,8,8a-
octahydronaphthalen-1-yl)methylene skeleton of this com-
pound, including total syntheses,7 biomimetic cyclizations,8
or hemisyntheses starting from polycyclic natural products.9
(7) (a) Tsujimori, H.; Bando, M.; Mori, K. Eur. J. Org. Chem. 2000,
297. (b) Okamura, W. H.; Peter, R.; Reischl, N. J. Am. Chem. Soc. 1985,
107, 1034. (c) Schmidt, C.; Chisti, N. H.; Breining, T. Synthesis 1982, 391.
(d) Arima, Y.; Kinoshita, M.; Akita, H. Tetrahedro: Asymmetry 2007, 18,
1701. (e) Akita, H.; Amano, H.; Kato, K.; Kinoshita, M. Tetrahedron:
Asymmetry 2004, 15, 725.
(8) (a) Tsangarakis, C.; Stratakis, M. AdV. Synth. Catal. 2005, 347, 1280.
(b) Polovinka, M. P.; Korchagina, D. V.; Gatilov, Y. V.; Bagrianskaya,
I. Y.; Barkhash, V. A.; Perutskii, V. B.; Ungur, N. D.; Vlad, P. F.;
Shcherbukhin, V. V.; Zefirov, N. S. J. Org. Chem. 1994, 59, 1509.
(9) (a) Kuchkova, K. I.; Aryku, A. N.; Barba, A. N.; Vlad, P. F. Chem.
Nat. Prod. 2007, 43, 412. (b) Barrero, A. F.; Alvarez-Manzaneda, E.;
Chahboun, R. Tetrahedron 1998, 54, 5635.
(10) (a) Kulcitki, V.; Ungur, N.; Gavagnin, M.; Carbone, M.; Cimino,
G. Eur. J. Org. Chem. 2005, 1816. (b) Dominguez, G.; Hueso-Rodriguez,
J. A.; de la Torre, M. C.; Rodriguez, B. Tetrahedron Lett. 1991, 32, 4765.
(11) Hagiwara, H.; Takeuchi, F.; Nozawa, M.; Hoshi, T.; Suzuki, T.
Tetrahedron 2004, 60, 1983.
(3) For recent reviews on the therapeutic potential of phosphoinositide-
3-kinase inhibitors, see: (a) Ward, S. G.; Sotsios, Y.; Dowden, J.; Bruce,
I.; Finan, P. Chem. Biol. 2003, 10, 207. (b) Ward, S. G.; Finan, P. Curr.
Opin. Pharmacol. 2003, 3, 426.
(4) Grube, A.; Assmann, M.; Lichte, E.; Sasse, F.; Pawlik, J. R.; Ko¨ck,
M. J. Nat. Prod. 2007, 70, 504.
(12) Abad, A.; Agullo´, C.; Castelblanque, L.; Cun˜at, A. C.; Navarro, I.;
Ramirez de Arellano, M. C. J. Org. Chem. 2000, 65, 4189.
(13) Tanimoto, H.; Oritani, T. Tetrahedron: Asymmetry 1996, 7, 1695.
(14) Alvarez-Manzaneda, E.; Chahboun, R.; Cabrera, E.; Alvarez, E.;
Haidour, A.; Ramos, J. M.; Alvarez-Manzaneda, R.; Hmammouchi, M.;
Es-Samti, H. Chem. Commun. 2009, 592.
(5) Mehta, G.; Likhite, N. S.; Kumar, C. S. A. Tetrahedron Lett. 2009,
50, 5260.
(6) Grube, A.; Assmann, M.; Lichte, E.; Sasse, F.; Pawlik, J. R.; Ko¨ck,
M. J. Nat. Prod 2007, 70, 504.
Org. Lett., Vol. 12, No. 20, 2010
4451

Treatment of this with the aryllithium derived from
bromide 12, which is straightforwardly prepared from the
very accessible sesamol,15 gave epoxy alcohols 7a,b as a
1:1 mixture, which was resolved after combined column
chromatography and crystallization. Oxidation of the mixture
of epoxy alcohols 7a,b with PDC gave ketone 13, which
after treatment with LiAlH4 in Et2O cooled at -78 °C and
further stirring at room temperature for 80 min gave
compounds 7a,b in a 5:1 ratio; further recrystallization
afforded pure stereoisomer 7a. The configuration on the C-11
of compounds 7a and 7b was established on the basis of
chemical evidence and by chemical correlation with com-
pounds 5a and 5b. Epoxy alcohol 7b underwent fast epoxide
ring-opening with complete regio- and stereoselectivity to
give the corresponding 11S diol 6b by refluxing with LiAlH4
in THF; under the same reaction conditions, epoxy alcohol
7a was slowly converted into the corresponding 11R diol
6a. This could be attributed to the steric hindrance of angular
methyl and the aromatic ring arising in the alkoxyaluminium
intermediate generated during the reduction of compound
7a.
of compound 5a in the presence of perchloric acid afforded
in excellent yield benzofuran 14, which possesses the
tetracyclic liphagane skeleton. Under the same reaction
conditions, the isomer 5b gave compound 14 in trace
amounts, together with an unresolvable mixture of com-
pounds. All attempts at converting compound 5b into its
epimer 5a, under different acid (e.g., concd HCl or concd
H2SO4 in refluxing dioxane) or basic conditions (e.g., DBU
in refluxing toluene or KOH in diglyme under reflux), were
unsuccessful.
Scheme 4. Direct Synthesis of Liphagane Precursor 14
Scheme 3. Construction of the “Liphagane” Skeleton
Once the liphagane precursor 14 was achieved, our efforts
were directed toward shortening the synthetic sequence, and
so one-step benzyl ether deprotection and B ring expansion
were investigated. With this purpose in mind, the rearrange-
ment of diol 6a promoted by perchloric acid was studied.
Unfortunately, the treatment of compound 6a with this acid
afforded a 1:1 mixture of epimers 5a,b. After perchloric acid
was ruled out, the utilization of mild acid conditions was
investigated. Diol 6a was transformed into ketone 5a in good
yield with complete stereospecificity by treatment with
Amberlyst A-15 in methanol. Interestingly, the dihydrox-
yphenol 15 underwent simultaneous rearrangement and
formation of the furan ring, directly affording the liphagane
intermediate 14, after cationic resin treatment (Scheme 4).
Next, hydrogenation in the presence of cationic resin was
investigated. The treatment of a mixture of ketone 5a,
palladium on carbon, and Amberlyst in methanol, under a
hydrogen atmosphere, gave in good yield the liphagane
precursor 14. Finally, the simultaneous rearrangement, benzyl
ether deprotection, and furan ring formation were tackled.
When a methanolic solution of diol 6a was treated with
palladium on carbon and cationic resin, under a hydrogen
atmosphere, the liphagane compound 14 was obtained
(Scheme 4); the hydroxyl phenol 16 was also obtained as a
minor constituent when the hydrogen pressure was increased
and the proportion of cationic resin reduced.
The next step was to address the construction of the
liphagane skeleton (scheme 3). The elaboration of the fused
6,7-ring system was achieved after the pinacol rearrangement
of diols 6a,b. Treatment of the 11R diol 6a with POCl3 and
pyridine in dichloromethane at -50 °C for 40 min gave in
87% yield the corresponding cycloheptanone 5a; the 11S diol
6b under these reaction conditions led to the epimeric ketone
5b. The C-10 configurations for compounds 5a and 5b were
established on the basis of NOE experiments. Hydrogenation
A possible mechanism for the direct formation of the
liphagane precursor 14 from diol 6a, via dihydroxy phenol
(15) Hitotsuyanagi, Y.; Ichihara, Y.; Takeya, K.; Itokawa, H. Tetrahe-
dron Lett. 1994, 35, 9401.
4452
Org. Lett., Vol. 12, No. 20, 2010

Scheme 5
.
Mechanism for the Formation of 14 and 16
Scheme 6. Synthesis of Liphagal (1) from Benzofuran 14
the same spectroscopic properties as reported in the literature.
The [R]D +17.9 [lit.2 +12.0] confirms the absolute stereo-
chemistry.
In summary, the first enantioselective synthesis of the
selective inhibitor of PI3 kinase R liphagal (1) is reported;
this allows us to establish absolute stereochemistry for this
marine metabolite. Key steps of the synthetic sequence are
the chemo- and stereoselective epoxidation of a homodrimane
diene and the one-step transformation of a dihydroxydrimane
phenol benzyl ether into the tetracyclic “liphagane” precursor,
which involves a stereospecific pinacol rearrangement, the
benzyl ether deprotection, the formation of a hemiketal, and
its subsequent dehydration.17
15, promoted by cationic resin, is depicted in Scheme 5. The
cycloheptanone derivative II could be generated after the
rearrangement of the hydroxy trienone I; further hemiket-
alization and dehydration would lead to the liphagane
intermediate 14. The hydroxyl phenol 16 would be formed
by reduction of the intermediate I, prior to the rearrangement
process. The lack of stereospecificity observed during the
perchloric acid promoted ring expansion can be attributed
to the formation of an intermediate benzyl cation.
Acknowledgment. We thank the Spanish Ministry of
Science and Innovation (Project No. CTQ2009-09932).
Finally, compound 14 was transformed into liphagal (1)
(Scheme 6). Treatment of this with BuLi in THF at -78 °C
followed by the addition of DMF and further reaction for
2 h gave aldehyde 17. Deprotection of the methylenedioxy
group was achieved utilizing a modification of the Imakura
procedure.16 A 6:1 mixture of phenylsulfides 18a and 18b
resulted when compound 17 was heated with PhSH and
K2CO3 in HMPA at 160 °C for 10 min. Refluxing a
methanolic solution of sulfides 18a,b in the presence of
catalytic concd HCl led to liphagal (1). This compound had
Note Added after ASAP Publication. Reference 17 was
added to the version reposted on June 17, 2010.
Supporting Information Available: Experimental details
and 1H NMR and 13C NMR spectra for compounds 1, 5a,b,
6a,b, 7a,b, 9-11, and 13-18. This material is available free
of charge via the Internet at http://pubs.acs.org.
OL101173W
(17) While this manuscript was in preparation, a similar approach to
liphagal was published, see: George, J. H.; Baldwin, J. E.; Adlington, R. M.
Org. Lett. 2010, 12, 2394.
(16) Imakura, Y.; Konishi, T.; Uchida, K.; Sakurai, H.; Kobayashi, S.;
Haruno, A.; Tajima, K.; Yamashita, S. Chem. Pharm. Bull. 1994, 42, 500.
Org. Lett., Vol. 12, No. 20, 2010
4453