Biomimetic Total Synthesis of Hyperjapones A–E and Hyperjaponols A and C

Article abstract of DOI:10.1002/anie.201606091

Hyperjapones A–E and hyperjaponols A–C are complex natural products of mixed aromatic polyketide and terpene biosynthetic origin that have recently been isolated from Hypericum japonicum. We have synthesized hyperjapones A–E using a biomimetic, oxidative

Full text of DOI:10.1002/anie.201606091

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201606091
German Edition: DOI: 10.1002/ange.201606091
Biomimetic Synthesis Very Important Paper
Biomimetic Total Synthesis of Hyperjapones A–E and
Hyperjaponols A and C
Hiu C. Lam, Justin T. J. Spence, and Jonathan H. George*
Abstract: Hyperjapones A–E and hyperjaponols A–C are
complex natural products of mixed aromatic polyketide and
terpene biosynthetic origin that have recently been isolated
from Hypericum japonicum. We have synthesized hyperjapo-
nes A–E using a biomimetic, oxidative hetero-Diels–Alder
reaction to couple together dearomatized acylphloroglucinol
and cyclic terpene natural products. Hyperjapone A is pro-
posed to be the biosynthetic precursor of hyperjaponol C
through a sequence of: 1) epoxidation; 2) acid-catalyzed
epoxide ring-opening; and 3) a concerted, asynchronous
alkene cyclization and 1,2-alkyl shift of a tertiary carbocation.
Chemical mimicry of this proposed biosynthetic sequence
allowed a concise total synthesis of hyperjaponol C to be
completed in which six carbon–carbon bonds, six stereocenters,
and three rings were constructed in just four steps.
a common 4-9-6-6 ring system. Further racemic natural
products, hyperjaponols A–C[2] (6–8), have also been isolated
from Hypericum japonicum. The most structurally intriguing
of these natural products is hyperjaponol C (8) because its
stereochemically complex, but racemic, trans-isodaucane
framework implies a highly predisposed, non-enzymatic
biosynthesis.
Our detailed proposal for the biosynthesis of hyperjapo-
ne A (1) and its conversion into hyperjaponols A–C (6–8) is
presented in Scheme 1. First, trimethylation of acylphloro-
glucinol 9 would give the dearomatized natural product
norflavesone (10).[3] Oxidation of 10 would give the a,b-
unsaturated ketone 12, which could be expected to exhibit
similar reactivity to an o-quinone methide.[4] It could there-
fore undergo a non-enzymatic, hetero-Diels–Alder reaction
with humulene (11) to give racemic hyperjapone A (1). The
D1,2 alkene of humulene was shown to be the most reactive
dieneophile in hetero-Diels–Alder reactions between humu-
lene and an o-quinone methide in both the biomimetic
synthesis of lucidene by Baldwin et al.[5] and the biomimetic
synthesis of guajadial B by Liu and co-workers.[6] The
biosynthesis of hyperjapone A (1) should therefore be
inherently regioselective with respect to humulene (11). The
hetero-Diels–Alder reaction would also be expected to be
regioselective with respect to the a,b-unsaturated ketone 12
as its most stable tautomer would be stabilized by an
intramolecular hydrogen bond (as shown in Scheme 1).
Hyperjapones B–E (2–5) could arise from a similar non-
enzymatic, oxidative hetero-Diels–Alder reaction between
a trimethylated acylphloroglucinol and the reactive trans D4,5
alkene of caryophyllene (18).[7]
H
yperjapones A–E[1] (Figure 1, 1–5) are a family of five
structurally related meroterpenoids isolated from Hypericum
japonicum, a plant used to treat hepatitis in traditional
Chinese medicine. Hyperjapone A is a racemic natural
product with an 11-6-6 tricyclic ring system. Hyperjapo-
nes B–E are enantiopure natural products which share
The X-ray crystal structure of hyperjapone A (1) shows
that it adopts a conformation with one face of the D8,9 alkene
exposed, while the other face is sterically hindered inside the
11-membered ring (Scheme 1).[1] Diastereoselective epoxida-
tion of the exposed face of the D8,9 alkene would therefore
give epoxide 13, which has a 1R*2R*8S*9S* relative config-
uration. A similarly regio- and diastereoselective result was
obtained in a recent diepoxidation of humulene (11) by Fujita
and co-workers, in which the structure of the diepoxide
product was elucidated using the crystalline sponge method.[8]
Acid-catalyzed ring opening of epoxide 13 would generate
tertiary carbocation 14, with the carbocation centered at the
C9 position. Deprotonation of 14 could then give either
hyperjaponol A (6) or hyperjaponol B (7). Alternatively,
a stereoselective cation–alkene cyclization of 14 could give
secondary carbocation 15, with the relative configuration of
the newly formed trans ring junction dictated by the orienta-
tion of the D4,5 alkene in the favored hyperjapone A
conformation. A stereospecific 1,2-alkyl shift of secondary
Figure 1. Hyperjapones A–E and hyperjaponols A–C.
[*] H. C. Lam, J. T. J. Spence, Dr. J. H. George
Department of Chemistry
University of Adelaide
Adelaide, SA 5005 (Australia)
E-mail: jonathan.george@adelaide.edu.au
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201606091.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Scheme 1. Proposed biosynthesis of hyperjaponol C from hyperjapone A through a concerted, asynchronous alkene cyclization/1,2-alkyl shift.
carbocation 15 would then generate tertiary carbocation 16,
and a final deprotonation would give hyperjaponol C (8).
Alternatively, it is perhaps more likely that carbocation 14
would directly rearrange to give carbocation 16 through
a concerted, asynchronous alkene cyclization/1,2-alkyl shift.
Similar concerted carbocation rearrangements in terpene
biosynthesis have been proposed and studied computationally
by Tantillo and co-workers.[9]
The primary aim of this project was to use the biosynthetic
hypothesis outlined in Scheme 1 as the blueprint for a diver-
gent, biomimetic synthesis of hyperjaponols A–C (6–8) from
hyperjapone A (1). Our synthesis of hyperjapone A began
with Friedel–Crafts acylation of phloroglucinol (17) with
isobutyryl chloride to give an acylphloroglucinol,[10] which was
dearomatized by means of trimethylation[11] to give norflave-
sone (10) in good yield over two steps (Scheme 2). Oxidation
of 10 with TEMPO and Ag2O in the presence of humulene
(11) then generated hyperjapone A (1) in 32% yield,
presumably through a hetero-Diels–Alder reaction between
the a,b-unsaturated ketone 12 generated in situ and humu-
lene (11). The use of TEMPO and Ag2O in tandem was the
result of extensive reaction optimization (see the Supporting
Scheme 2. Total synthesis of hyperjapone A and hyperjaponols A
Information for full details). Ag2O has previously been used
for the oxidative generation of o-quinone methides,[12] but not
in the oxidative formation of a,b-unsaturated ketones. We
have previously used a combination of TEMPO and PhI-
(OAc)2 to induce a similar oxidation to generate an a,b-
unsaturated ketone that underwent a 6p-electrocyclization in
our biomimetic synthesis of hyperguinone B.[13]
and C. TEMPO=(2,2,6,6-tetramethylpiperidin-1-yl)oxyl; m-CPBA=3-
chloroperoxy-benzoic acid; p-TsOH=p-toluenesulfonic acid.
japonol C (8). As expected from molecular modelling studies
and Fujitaꢀs humulene epoxidation work,[8] treatment of
1 with m-CPBA generated epoxide 13 as a single diastereo-
mer in 76% yield through diastereoselective oxidation of the
D8,9 alkene. We were then delighted to observe that acid-
With a concise synthesis of hyperjapone A (1) established,
we next focused on its biomimetic conversion into hyper-
2
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
catalyzed rearrangement of 13 using p-TsOH in CH2Cl2 gave
hyperjaponol C (8) in 43% yield, presumably by the con-
certed, asynchronous alkene cyclization/1,2-alkyl shift mech-
anism outlined in Scheme 1. The use of AgNO3/SiO2 chro-
matography was essential in the purification of 8 from trace
reaction by-products.[14] The synthesis of 8 could be stream-
lined by combining the epoxidation and acid-catalyzed
rearrangement steps into a one-pot procedure, which directly
converts 1 into 8 in 34% yield.[15] This operationally simple
reaction diastereoselectively generates four stereocenters of
the unusual trans-isodaucane skeleton of 8. Almost all
isodaucane terpenoid natural products previously isolated
possess a cis-fused ring junction that is likely to be biosyn-
thesized through cyclization of a germacrene D derivative.[16]
Several acid-catalyzed rearrangements of humulene[17] and
humulene-8,9-epoxide[18] that are similar to the conversion of
1 into 8 have been previously reported. However, these
rearrangements are generally unselective, giving rise to
complex mixtures of products. Furthermore, our synthesis of
8 is the first time that this rearrangement mode of a humulene
derivative has been used in a natural product synthesis, and it
is the first time that it has been proposed to occur in
a biosynthetic pathway.
Scheme 4. Total synthesis of hyperjapones C and E.
The total syntheses of hyperjapones C (3) and E (5) was
achieved in a similar fashion (Scheme 4). Friedel–Crafts
acylation of phloroglucinol (17) with (S)-2-methylbutyryl
chloride[21] followed by trimethylation gave the dearomatized
natural product norisoleptospermone (19). Exposure of 19 to
TEMPO/Ag2O in the presence of caryophyllene (18) gave
a 2.5:1 mixture of hyperjapone C (3) and hyperjapone E (5) in
61% combined yield. Again, the product ratio of the
oxidative hetero-Diels–Alder reaction approximately corre-
lates to the ratio of the ba and bb conformations of 18.
Additionally, we synthesized 20 and 21 as an inseparable
1:1 mixture of diastereomers by treatment of 19 with
TEMPO/Ag2O in the presence of humulene (11). Given the
probable biosynthesis of hyperjapones A–E in Hypericum
japoniucm through non-enzymatic hetero-Diels–Alder reac-
tions, it is highly likely that 20 and 21 are “undiscovered
natural products”. Samples of 20 and 21, and also hyper-
japone A epoxide (13), have therefore been distributed to
isolation chemists in order to accelerate their discovery in
nature.
Conversion of epoxide 13 into hyperjaponol A (6) was
=
achieved in 59% yield upon treatment with (NC)2C C(CN)2
and LiBr in acetone.[19] Despite extensive screening of further
Lewis and protic acids, the formation of hyperjaponol B (7)
from 13 has not yet been observed.
The secondary aim of this project was to synthesize
hyperjapones B–E using an oxidative hetero-Diels–Alder
reaction to couple caryophyllene (18) with trimethylated
acylphloroglucinol natural products. Thus, treatment of nor-
flavesone (10) with TEMPO/Ag2O in the presence of
caryophyllene (18) gave a 2.5:1 mixture of hyperjapone B
(2) and hyperjapone D (4) in 60% combined yield
(Scheme 3). 18 exists as a 3:1 mixture of ba and bb
In conclusion, we have used a biomimetic oxidative
hetero-Diels–Alder reaction to synthesize hyperjapones A–
E (1–5). This strategy allows for efficient coupling of
dearomatized, trimethylated acylphloroglucinol natural prod-
ucts to the reactive trans alkenes of cyclic terpenes. Hyper-
japone A (1) was converted into hyperjaponol A (6) and
hyperjaponol C (8) through acid-catalyzed rearrangement of
an intermediate epoxide. The four-step synthesis of hyper-
japonol C (8) from simple starting materials involves the
construction of six carbon–carbon bonds, six stereocenters,
and three rings, and is thus a good example of the use of
a biomimetic synthetic approach to rapidly generate molec-
ular complexity. More generally, this work shows that
a biomimetic approach to synthesis can lead to both naturally
divergent strategies (nine natural products and three pro-
posed “undiscovered natural products” have been synthe-
sized, all for the first time) and naturally efficient syntheses
Scheme 3. Total synthesis of hyperjapones B and D.
conformations in solution.[20] 2 is formed by the cycloaddition
of the a,b-unsaturated ketone 12 generated in situ to the
reactive trans D4,5 alkene of the more abundant ba conforma-
tion, while 4 is generated from addition to the less abundant
bb conformation.[7] In their isolation paper, Xu et al. reported
the separation of 2 and 4 using preparative HPLC.[1] How-
ever, we also found that 2 could be purified from 4 by means
of selective crystallization from MeOH.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
2011, 13, 5318; c) K. Takao, S. Noguchi, S. Sakamoto, M. Kimura,
K. Yoshida, K. Tadano, J. Am. Chem. Soc. 2015, 137, 15971.
[8] N. Zigon, M. Hoshino, S. Yoshioka, Y. Inokuma, M. Fujita,
Angew. Chem. Int. Ed. 2015, 54, 9033; Angew. Chem. 2015, 127,
9161.
(all routes are protecting-group-free, with excellent step,
redox, and pot economy).
Acknowledgements
[9] a) D. J. Tantillo, Chem. Soc. Rev. 2010, 39, 2847; b) D. J. Tantillo,
Nat. Prod. Rev. 2011, 28, 1035.
We thank the Australian Research Council for financial
support in the form of a Discovery Project (DP160103393).
[10] L. Crombie, R. C. F. Jones, C. Palmer, J. Chem. Soc. Perkin
Trans. 1 1987, 317.
[11] N. T. Nguyen, V. C. Pham, M. Litaudon, F. Guꢁritte, B. Bodo,
V. T. Nguyen, V. H. Nguyen, Tetrahedron 2009, 65, 7171.
[12] D. Liao, H. Li, X. Lei, Org. Lett. 2012, 14, 18.
[13] J. H. George, M. D. Hesse, J. E. Baldwin, R. M. Adlington, Org.
Lett. 2010, 12, 3532.
Keywords: biomimetic synthesis · carbocation rearrangements ·
cascade reactions · natural products · terpenoids
[14] a) C. M. Williams, L. N. Mander, Tetrahedron 2001, 57, 425;
b) L. N. Mander, C. M. Williams, Tetrahedron 2016, 72, 1133.
[15] For a similarly telescoped one-pot reaction involving epoxida-
tion followed by acid-catalyzed rearrangement, see: A. W.
Markwell-Heys, K. K. W. Kuan, J. H. George, Org. Lett. 2015,
17, 4228.
[16] N. Bꢂlow, W. A. Kçnig, Phytochemistry 2000, 55, 141.
[17] a) Y. Naya, Y. Hirose, Chem. Lett. 1973, 2, 133; b) Y. Naya, Y.
Hirose, Chem. Lett. 1973, 2, 727; c) D. Baines, J. Forrester, W.
Parker, J. Chem. Soc. Perkin Trans. 1 1974, 1598; d) W. G.
Dauben, J. P. Hubbell, N. D. Vietmeyer, J. Org. Chem. 1975, 40,
479.
[18] I. Bryson, J. S. Roberts, A. Sattar, Tetrahedron Lett. 1980, 21, 201.
[19] I. G. Collado, J. R. Hanson, A. J. Macꢃas-Sꢄnchez, Tetrahedron
1996, 52, 7961.
[20] M. Clericuzio, G. Alagona, C. Ghio, L. Toma, J. Org. Chem.
2000, 65, 6910.
[1] X.-W. Yang, Y.-P. Li, J. Su., W.-G. Ma, G. Xu, Org. Lett. 2016, 18,
1876.
[2] L. Hu, Y. Zhang, H. Zhu, J. Liu, H. Li, X.-N. Li, W. Sun, J. Zeng,
Y. Xue, Y. Zhang, Org. Lett. 2016, 18, 2272.
[3] D. P. Killeen, L. Larsen, F. E. Dayan, K. C. Gordon, N. B. Perry,
J. W. van Klink, J. Nat. Prod. 2016, 79, 564.
[4] For reviews of o-quinone methide chemistry, see: a) R. W.
Van De Water, T. R. R. Pettus, Tetrahedron 2002, 58, 5367;
b) N. J. Willis, C. D. Bray, Chem. Eur. J. 2012, 18, 9160; c) W.-J.
Bai, J. G. David, Z.-G. Feng, M. G. Weaver, K.-L. Wu, T. R. R.
Pettus, Acc. Chem. Res. 2014, 47, 3655.
[5] a) R. M. Adlington, J. E. Baldwin, G. J. Pritchard, A. J. Williams,
D. J. Watkin, Org. Lett. 1999, 1, 1937; b) R. Rodriguez, J. E.
Moses, R. M. Adlington, J. E. Baldwin, Org. Biomol. Chem.
2005, 3, 3488.
[21] M. J. Begley, L. Crombie, R. C. F. Jones, C. J. Palmer, J. Chem.
Soc. Perkin Trans. 1 1987, 353.
[6] Y. Gao, G.-Q. Wang, K. Wei, P. Hai, F. Wang, J.-K. Liu, Org. Lett.
2012, 14, 5936.
[7] For examples of biomimetic hetero-Diels–Alder reactions
involving caryophyllene, see: a) A. L. Lawrence, R. M. Adling-
ton, J. E. Baldwin, V. Lee, J. A. Kershaw, A. L. Thompson, Org.
Lett. 2010, 12, 1676; b) J. T. J. Spence, J. H. George, Org. Lett.
Received: June 23, 2016
Published online: && &&, &&&&
4
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Biomimetic Synthesis
H. C. Lam, J. T. J. Spence,
J. H. George*
&&&& — &&&&
Biomimetic Total Synthesis of
Hyperjapones A–E and Hyperjaponols A
and C
Chemical mimicry: A biomimetic oxida-
tive hetero-Diels–Alder reaction was
employed in the total syntheses of
hyperjapones A–E. Chemical mimicry of
the proposed biosynthetic pathway for
the conversion of hyperjapone A into
hyperjaponol C allowed a concise total
synthesis of hyperjaponol C to becom-
pleted in just four steps.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!