Enantioselective total synthesis of dysidavarone A, a novel sesquiterpenoid quinone from the marine sponge dysidea avara

Article abstract of DOI:10.1002/chem.201304809

Dysidavaronea A, a structurally unprecedented sesquiterpenoid quinone, was synthesized in 30 % overall yield in a longest liner sequence of 13 steps from commercially available o-vanillin. A highly strained and bridged eight-membered carbocyclic core was established by the C7-C21 carbon bond formation through a copper enolate mediated Michael addition to the internal quinone ring. The power of a copper enolate! Dysidavaronea A, a structurally unprecedented sesquiterpenoid quinone, was synthesized in 30 % overall yield in a longest linear sequence of 13 steps from commercially available o-vanillin (see scheme; TBS=tert-butyldimethylsilyl). A highly strained and bridged eight-membered carbocyclic core was established by the C7-C21 carbon bond formation through a copper enolate mediated Michael addition to the internal quinone ring. Copyright

Full text of DOI:10.1002/chem.201304809

DOI: 10.1002/chem.201304809
Communication
&
Natural Product Synthesis
Enantioselective Total Synthesis of Dysidavarone A, a Novel
Sesquiterpenoid Quinone from the Marine Sponge Dysidea avara
Yurie Fukui, Koichi Narita, and Tadashi Katoh*[a]
Dedicated to Professor Philip D. Magnus FRS on the occasion of his 70th birthday
Abstract: Dysidavarone A, a structurally unprecedented
sesquiterpenoid quinone, was synthesized in 30% overall
yield in a longest liner sequence of 13 steps from com-
mercially available o-vanillin.
A highly strained and
bridged eight-membered carbocyclic core was established
by the C7ꢀC21 carbon bond formation through a copper
enolate mediated Michael addition to the internal quinone
ring.
In 2012, Lin et al. reported the isolation and structural elucida-
tion of four novel sesquiterpenoid quinones, dysidavarones A–
D (1–4, Figure 1), from the marine sponge, Dysidea avara, col-
lected along the coast of Yongxing Island in the South China
Sea.[1] These marine natural products were found to have an
unprecedented tetracyclic dysidavarane carbon skeleton (see
structure 5) with a highly strained and bridged eight-mem-
bered carbocycle.[1] Lin et al. proposed that this unusual tetra-
cyclic carbon skeleton might be produced biogenetically from
well-known drimane-type sesquiterpenoid quinones, avarone
(6) and neoavarone (7), by the C7ꢀC21 bond formation
through an intramolecular Prins-type reaction.[1] Compounds
1 and 4 have shown inhibitory activity against protein tyrosine
phosphatase 1B (PTP1B) with IC50 values of 9.98 and 21.6 mm,
respectively.[1] PTP1B is a major negative regulator in both insu-
lin and leptin signalling pathways, and it acts as a positive reg-
ulator in the tumorigenesis and progression of cancers.[2] It has
also been shown that 1 and 4 exhibit antiproliferative activities
against several human cancer cell lines in the micromolar
range (IC50 =11.6–28.8 mm).[1] More recently, 1 has been dem-
onstrated to show potent inhibitory effects against Gram-posi-
tive bacteria, in particular against various Staphylococci
(MIC50 =0.2–9.9 mgmLꢀ1).[3] Therefore, we anticipated that dysi-
davarones may be promising candidates or provide new leads
for the treatment of pathogenic states such as diabetes, obesi-
Figure 1. Structures of dysidavarones A–D (1–4), dysidavarane skeleton (5),
avarone (6) and neoavarone (7).
ty, cancer and infections. However, further biological studies of
these natural products are severely restricted by the scarcity of
samples from marine sponge Dysidea species (0.0002–0.0004%
yields upon isolation). Consequently, developing methods for
synthesis of 1–4 is desirable and worthwhile from the view-
point of medicinal chemistry and pharmaceuticals.
Their unique structural features, attractive biological activi-
ties and limited availability from natural resources have made
1–4 exceptionally intriguing and timely targets for total syn-
thesis. Recently, while our synthetic studies of 1–4 were in
progress, Menche et al. reported the first and elegant total syn-
thesis of 1, which was rendered in a longest liner sequence of
10 steps with 11% overall yield.[3] Their synthetic route is out-
lined in Scheme 1, where the strategic Pd-catalyzed intramo-
lecular a-arylation[4] of appropriately elaborated ketone 10,
which was prepared from known (+)-enone 8[5] and aryl bro-
mide 9, to construct the requisite tetracyclic structure 11 was
the key step (10!11, 66% yield).[3] Herein, we describe our
total synthesis of 1 using an alternative strategy that involves
an advantageous intramolecular Michael addition to construct
the requisite tetracyclic core structure.
[a] Y. Fukui, K. Narita, Prof. Dr. T. Katoh
Laboratory of Synthetic and Medicinal Chemistry
Faculty of Pharmaceutical Sciences
Tohoku Pharmaceutical University
4-4-1 Komatsushima, Aoba-ku, Sendai, 981-8558 (Japan)
Fax: (+81)22-727-0135
Our retrosynthetic analysis of dysidavarone A (1) is illustrated
in Scheme 2. We envisioned that the requisite tetracyclic struc-
ture 13 could be established from appropriately functionalized
E-mail: katoh@tohoku-pharm.ac.jp
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304809.
Chem. Eur. J. 2014, 20, 2436 – 2439
2436
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Scheme 3. Synthesis of intermediate 15. Reaction conditions: a) TBSCl, imi-
dazole, CH2Cl2, RT, 12 h, 97%; b) NaBH4, THF/H2O, 08C, 30 min, 99%; c) PBr3,
CH2Cl2, 08C to RT, 5 min, 98%. TBS=tert-butyldimethylsilyl.
from o-vanillin (16) (Scheme 3). After protection of the hydroxy
group in 16 (97% yield), the formyl group in the resulting tert-
butyldimethylsilyl (TBS) ether 17 was reduced with NaBH4 to
provide benzylic alcohol 18 in quantitative yield. Subsequent
bromination of 18 (PBr3, CH2Cl2, 08C to RT) afforded the de-
sired intermediate 15 in 98% yield.
Scheme 1. Outline of the total synthesis of dysidavarone A (1) by Menche
et al.
Having synthesized intermediate 15, we approached the
synthesis of intermediate 14, a precursor of the key cyclization
reaction, as shown in Scheme 4. The critical reductive alkyla-
Scheme 2. Retrosynthetic analysis of dysidavarone A (1).
Scheme 4. Synthesis of intermediate 14. Reaction conditions: a) Li, NH3/THF,
ꢀ78 to ꢀ308C, 45 min; isoprene, 15, ꢀ78 to ꢀ308C, 2 h, 81%; b) TBAF, THF,
RT, 30 min, 98%; c) O2 (1 atm), salcomine, MeCN, RT, 15 h, 86%. TBAF=tetra-
n-butylammonium fluoride, salcomine=N,N’-bis(salicylidene)ethylenediami-
nocobalt(II).
decalone 14 by the C7ꢀC21 bond formation through a Michael
addition of the C8 carbonyl derived enolate to the internal qui-
none ring,[6] followed by in situ air oxidation of the resulting
hydroquinone. To the best of our knowledge, this type of cycli-
zation to construct a highly strained and bridged eight-mem-
bered carbocycle fused with a quinone ring has not been pre-
viously reported; thus, this approach posed a considerable
challenge from a synthetic viewpoint.[7] The cyclization product
13 could be converted into the target molecule 1 via the ad-
vanced intermediate 12 by deprotection and functional-group
manipulation or vice versa. The cyclization precursor 14, in
turn, was to be prepared by the stereocontrolled reductive al-
kylation of known optically active enone 8[5] and benzyl bro-
mide 15, which is accessible from commercially available inex-
pensive o-vanillin (16), under Birch conditions. This devised
scheme involves the substitution reaction of the alkoxy group
in the final stage of the synthesis (see 12!1); therefore, this
method is applicable to the synthesis of structural analogues
of 1 possessing various alkoxy groups on the quinone ring.
We first pursued the synthesis of intermediate 15, the cou-
pling partner of the decalin portion (see structure 8), starting
tion of enone 8[5] (>99% ee) with 15 under Birch conditions[8]
proceeded smoothly and cleanly to give the expected cou-
pling product 19 in 81% yield as the single diastereomer. De-
protection of the TBS group in 19 resulted in the formation of
hemiacetal 20 in 98% yield. The stereochemistry at the C8 po-
sition in 20 was confirmed by NOESY experiment (see the Sup-
porting Information). To construct the quinone system directly,
20 was allowed to react with molecular oxygen (O2 balloon) in
the presence of salcomine, that is, N,N’-bis(salicylidene)ethyle-
nediaminocobalt(II),[9] in acetonitrile at ambient temperature
for 15 h, thus producing the desired quinone 14 in high yield
(86%).
After obtaining intermediate 14, we next investigated the
key cyclization reaction to construct the requisite dysidavarane
skeleton (14!I!II!13; see scheme above Table 1). The se-
quence involved Michael addition of the metallic enolate I[10,11]
to the internal quinone ring and subsequent air oxidation of
Chem. Eur. J. 2014, 20, 2436 – 2439
www.chemeurj.org
2437
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 1. Cyclization of 14 leading to dysidavarane skeleton 13.
Entry
1
Conditions[a]
Yield[b] [%]
dec.[c]
LiNiPr2 (3 equiv),
THF (1 mm), ꢀ78 to ꢀ408C
2
3
4
5
LiN(SiMe3)2 (5 equiv),
THF (1 mm), ꢀ408C to RT, 48 h
28
Scheme 5. Completion of the total synthesis of dysidavarone A (1). Reaction
conditions: a) 3m HCl, THF, RT, 62%; b) Ph3P+CH3Brꢀ, tBuOK, benzene, RT to
reflux, decomposition; c) 30% Na2S2O4, Me2SO4, 30% KOH, nBu4NBr, THF, RT,
80%; d) 3m HCl, THF, 408C, 98%; e) Ph3P+CH3Brꢀ, tBuOK, benzene, reflux,
95%; f) 3m HCl, THF, reflux, 87%; g) CAN, MeCN/H2O, ꢀ78C, 97%; h) DBU,
EtOH, RT, 90%. CAN=ceric ammonium nitrate, DBU=1,8-diazabicyclo-
[5.4.0]undec-7-ene.
NaN(SiMe3)2 (5 equiv),
THF (1 mm), ꢀ408C to RT, 24 h
~5
KN(SiMe3)2 (3 equiv),
THF (1 mm), ꢀ78 to ꢀ408C
dec.[c]
84
LiN(SiMe3)2 (5 equiv),
CuBr·SMe2 (2 equiv),
THF (1 mm), ꢀ408C to RT, 48 h
Michael addition reaction of a copper enolate with a quinone
to construct a highly strained medium-sized carbocycle.
The final route that led to the completion of the total syn-
thesis of 1 is shown in Scheme 5. Initial attempts to realize the
direct conversion of diketone 21, which was prepared from 13
by deprotection of the ethylene acetal moiety, into bis-exo-
olefin 22 under Wittig methylenation conditions[13] (Ph3P+
CH3Brꢀ, tBuOK, benzene, RT to reflux) were unsuccessful and
resulted in decomposition of material. This result is likely due
to the presence of the sensitive quinone function in 21. There-
fore, we decided to mask the reactive quinone system in the
form of a p-dimethoxybenzene derivative.[14] Thus, methoxyqui-
none 13 was converted into the corresponding trimethoxyben-
zene 23 (80% yield) by reductive methylation (30% Na2S2O4,
Me2SO4, 30% KOH, nBu4NBr, THF, RT). After deprotection of the
ethylene acetal moiety in 23 (98% yield), the resulting dike-
tone 24 was subjected to two-fold Wittig methylenation to
produce bis-exo-olefin 25 in high yield (95%). The subsequent
site-selective isomerization of the C4 exo olefinic double bond
in 25 (3m HCl, THF, reflux)[3] furnished the corresponding endo
olefin 26 in 87% yield. The quinone system was then regener-
ated efficiently by treating 26 with ceric ammonium nitrate
(CAN) in acetonitrile/water at low temperature (ꢀ78C), giving
rise to the desired methoxyquinone 12 in 97% yield. Finally,
exposure of 12 to ethanol in the presence of 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU) at ambient temperature resulted in
the formation of the targeted dysidavarone A (1) in 90%
[a] All reactions were carried out in the presence of air. [b] Yield upon iso-
lation. [c] dec.=Decomposition.
the resulting intermediate II[12] during the cyclization reaction.
To achieve cyclization, initial screening was carried out using
several metal amide reagents, such as LiNiPr2, LiN(SiMe3)2, NaN-
(SiMe3)2 and KN(SiMe3)2 in
a THF solvent (entries 1–4,
Table 1).[10,11] Among these reagents, LiN(SiMe3)2 proved to be
superior from the viewpoint of reaction cleanliness (entry 2,
Table 1); however, the yield of the desired cyclized product 13
was not satisfactory (28%). In the cases of LiNiPr2 and KN-
(SiMe3)2, the reaction was not clean, and the complete decom-
position of material was observed (entries 1 and 4, Table 1).
When NaN(SiMe3)2 was used, only a trace amount of 13 (~5%
yield) was produced and a large amount of the starting materi-
al 14 (90–95% yield) was recovered (entry 3, Table 1). After re-
peated attempts, we found that the addition of CuBr·SMe2 was
extremely effective and reliable for this cyclization event
(entry 5, Table 1). Thus, reaction of 14 with LiN(SiMe3)2
(5 equiv) in the presence of CuBr·SMe2 (2 equiv) in a dilute THF
solution (1 mm) from ꢀ408C to RT over the course of 48 h, re-
sulted in the formation of 13 in high yield (84%). In this reac-
tion, we believed that copper enolate I (M=Cu) was generated
in situ[10] and that this intermediate species promoted the in-
tramolecular Michael addition reaction. Thus, the expected cyc-
lization reaction proceeded in a clean and efficient way. To the
best of our knowledge, this is the first successful example of
1
yield.[15] The spectroscopic properties (IR, H, and 13C NMR spec-
troscopy, and high-resolution mass spectrometry) of synthetic
1 were identical to those of natural 1.[1] The optical rotation of
Chem. Eur. J. 2014, 20, 2436 – 2439
www.chemeurj.org
2438
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
synthetic 1 , [a]2D5 = +105.6 (c=1.00, MeOH), matched that re-
ported previously, natural 1: [a]2D5 = +30 (c=1.00, MeOH),[1, 16]
synthetic 1: [a]2D5 = +125 (c=1.00, MeOH).[3]
[7] Total synthesis of strained terpenoid natural products has been recently
highlighted, see: K. J. Hale, Org. Lett. 2013, 15, 3181–3198.
[8] This type of reductive alkylation, originally explored by Stork et al. in
the 1960s (a representative study; see: a) G. Stork, P. Rosen, N. Gold-
man, R. V. Coombs, J. Tsuji, J. Am. Chem. Soc. 1965, 87, 275–286), has
been widely used as a critical step in the total synthesis of marine ses-
quiterpenoidquinones and hydroquinones. For examples; see: b) J. Sa-
kurai, T. Oguchi, K. Watanabe, H. Abe, S. Kanno, M. Ishikawa, T. Katoh,
Chem. Eur. J. 2008, 14, 829–837; c) S. Poigny, M. Guyot, M. Samadi, J.
Org. Chem. 1998, 63, 5890–5894; d) S. D. Bruner, H. S. Radeke, J. A. Tal-
larico, M. L. Snapper, J. Org. Chem. 1995, 60, 1114–1115.
[9] Salcomine oxidation has been used for the conversion of phenols into
quinones; for examples, see: a) K. Narita, K. Nakamura, Y. Abe, T. Katoh,
Eur. J. Org. Chem. 2011, 4985–4988; b) X. W. Liao, W. Liu, W. F. Dong,
B. H. Guan, S. Z. Chen, A. Z. Liu, Tetrahedron 2009, 65, 5709–5715; c) S.
Akai, K. Kakiguchi, Y. Nakamura, I. Kuriwaki, T. Dohi, S. Harada, O. Kubo,
N. Morita, Y. Kita, Angew. Chem. 2007, 119, 7602–7605; Angew. Chem.
Int. Ed. 2007, 46, 7458–7461; d) T. Katoh, M. Nakatani, S. Shikita, R.
Sampe, A. Ishiwata, M. Nakamura, S. Terashima, Org. Lett. 2001, 3, 2701;
e) N. Saito, Y. Obara, T. Aihara, S. Harada, Y. Shida, A. Kubo, Tetrahedron
1994, 50, 3915–3928; f) K. Yoshida, S. Nakajima, T. Ohnuma, Y. Ban, M.
Shibasaki, K. Aoe, T. Date, J. Org. Chem. 1988, 53, 5355–5359; g) T. Wa-
kamatsu, T. Nishi, T. Ohnuma, Y. Ban, Synth. Commun. 1984, 14, 1167–
1173.
In summary, we have accomplished the enantioselective
total synthesis of dysidavarone A (1) in 30% overall yield in 13
steps from the starting material o-vanillin (16). The most crucial
step of the synthesis involved the formation of the eight-mem-
bered carbocyclic core by the strategic intramolecular Michael
addition/in situ air oxidation to establish the tetracyclic dysida-
varane skeleton (14!13, Table 1). On the basis of this study,
we are currently synthesizing additional analogues of 1 (for ex-
ample, analogues possessing a variety of alkoxy groups on the
quinone ring) with the aim of exploring its structure–activity
relationships. In addition, further investigations to identify ad-
ditional action mechanisms of 1 using the synthetic sample are
in progress in our laboratories.
Acknowledgements
[10] For a review on the formation of metallic enolates, see: H. B. Mekelbur-
ger, C. S. Wilcox, in Comprehensive Organic Synthesis Vol. 2 (Eds.: B. M.
Trost, I. Fleming), Pergamon, Oxford, 1991, pp. 99–131.
[11] Intermolecular Michael additions of ketone enolates to quinones have
been reported, see: a) N. V. Sastry Mudiganti, S. Claessens, N. D. Kimpe,
Tetrahedron 2009, 65, 1716–1723; b) W.-B. Kang, S. Nan’ya, T. Toru, Y.
Ueno, Chem. Lett. 1988, 1415–1418; c) T. Mukaiyama, N. Iwasawa, T.
Yura, R. S. J. Clark, Tetrahedron 1987, 43, 5003–5017.
This study was supported by a Grant-in-Aid for the Strategic
Research Foundation Program at Private Universities (2010–
2014) from the Ministry of Education, Culture, Sports, Science
and Technology, Japan (MEXT) and a Grant-in-Aid for Scientific
Research (C) (No. 24590017) from MEXT.
[12] The highly electron-rich hydroquinone moiety present in intermediate
II was found to be fairly unstable, and the autoxidation was observed
during the cyclization reaction.
Keywords: copper enolate
addition · sesquiterpenoids · total synthesis
· dysidavarone A · Michael
[13] L. Fitjer, U. Quabeck, Synth. Commun. 1985, 15, 855–864.
[14] C. L. L. Chai, J. A. Elix, F. K. Moore, J. Org. Chem. 2006, 71, 992–1001.
[15] We initially suspected that dysidavarone A (1) was the artefact derived
from 12 during the extraction operation by use of ethanol, because the
ethoxy substituent on the quinone ring in 1 is unusual in the naturally
occurring sesquiterpenoid quinones. In order to clear this suspicion, 12
was exposed to ethanol under neutral conditions (without basic and
acidic reagents) at ambient temperature over 24 h, resulting in no pro-
duction of 1. This result verified that 1 is not the artefact but the natu-
ral product itself.
[1] W.-H. Jiao, X.-J. Huang, J.-S. Yang, F. Yang, S.-J. Piao, H. Gao, J. Li, W.-C.
Ye, X.-S. Yao, W.-S. Chen, H.-W. Lin, Org. Lett. 2012, 14, 202–205.
[2] For reviews on PTP1B as a novel therapeutic target, see: a) S. Lee, Q.
Wang, Med. Res. Rev. 2007, 27, 553–573; b) S. Zhang, Z.-Y. Zhang, Drug.
Discov. Today 2007, 12, 373–381.
[3] B. Schmalzbauer, J. Herrmann, R. Mꢁller, D. Menche, Org. Lett. 2013, 15,
964–967.
[4] J. M. Fox, X. Huang, A. Chieffi, S. L. Buchwald, J. Am. Chem. Soc. 2000,
122, 1360–1370.
[5] H. Hagiwara, H. Uda, J. Org. Chem. 1988, 53, 2308–2311.
[6] This endo-type Michael addition is quite unique in that only one exam-
ple has appeared in the literature. Danishefsky et al. reported an intra-
molecular Michael addition of a ketone enolate to a quinone ring for
the construction of a benzo[d]anthracene core structure, see: D. M.
Gordon, S. J. Danishefsky, J. Org. Chem. 1992, 57, 7052–7055.
[16] It has been pointed out that the optical rotation of +30 may be
caused by a lower purity of the natural sample (see Ref. [3]).
Received: December 9, 2013
Published online on January 30, 2014
Chem. Eur. J. 2014, 20, 2436 – 2439
www.chemeurj.org
2439
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim