Biomimetic studies towards the cardinalins: Synthesis of (+)-ventiloquinone L and an unusual dimerisation

Article abstract of DOI:10.1039/b905077a

Studies towards the biomimetic synthesis of cardinalin 3 are described. Despite the successful enantioselective synthesis of the monomeric pyranonaphthoquinone ventiloquinone L, it subsequently failed to undergo a proposed biomimetic homodimerisation to c

Full text of DOI:10.1039/b905077a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Biomimetic studies towards the cardinalins: synthesis of (+)-ventiloquinone
L and an unusual dimerisation†
Jonathan Sperry, Jimmy J. P. Sejberg, Frank M. Stiemke and Margaret A. Brimble*
Received 12th March 2009, Accepted 6th April 2009
First published as an Advance Article on the web 28th April 2009
DOI: 10.1039/b905077a
Studies towards the biomimetic synthesis of cardinalin 3 are described. Despite the successful
enantioselective synthesis of the monomeric pyranonaphthoquinone ventiloquinone L, it subsequently
failed to undergo a proposed biomimetic homodimerisation to cardinalin 3 using a range of oxidants.
However, treatment of a related naphthopyran with cerium ammonium nitrate (CAN) facilitated a
tandem biaryl bond formation–oxidative demethylation sequence furnishing a dimeric
pyranonaphthoquinone that had exclusively dimerised at C6. The nature of this unusual sequence is
discussed and the product subsequently converted to the C6 regioisomer of cardinalin 3.
the monomeric pyranonaphthoquinone eleutherin,6 we envisaged
Introduction
ventiloquinone L 2 could be prepared using a Hauser–Kraus
annulation7 between cyanophthalide 3 and (-)-enone 4 (Scheme 1).
Cardinalin
3 (1) is a member of a family of dimeric
pyranonaphthoquinones1 contained in the deep-red ethanolic
extract of the fresh fruit bodies of the New Zealand toadstool
Dermocybe cardinalis.2 Its interesting structure, in particular the
tetra-substituted C8–C8¢ biaryl bond that results in 1 existing as
a single (aS)-atropisomer, combined with its significant cytotoxic
properties (IC50 = 0.47 mg mL-1 against P388 murine leukaemia
cell line) renders it an attractive target for synthesis (Fig. 1).
Fig. 1 Cardinalin 3.
Our previous synthetic efforts towards 1 have resulted in
the successful enantioselective synthesis of the dimeric pyra-
nonaphthoquinone core using a late stage homocoupling of an
aryltriflate.3 Additionally, de Koning et al. have reported a racemic
synthesis of 1 using a bidirectional approach in which the biaryl
bond was installed at an early stage of the synthesis.4
Herein we report our initial studies based on a revised approach
that hinges on the late stage biomimetic coupling of the monomer
of cardinalin 3 1, ventiloquinone L 2, itself a natural product
isolated from the root bark of Ventilago goughii.5 It was hoped
that if ventiloquinone L 2 could be made in an enantiopure
fashion, the chirality present in 2 may exert some induction
during the dimerisation step leading to atroposelective biaryl
formation. Adopting a route analogous to our synthesis of
Scheme 1 Retrosynthesis of 1.
Results and discussion
Thus, our initial thoughts turned to the enantioselective synthesis
of ventiloquinone L 2. De Koning et al. have reported a racemic
synthesis of 28 but to date only a single enantioselective synthesis
of 2 exists in the literature, a 10 step procedure employing the
Diels–Alder addition of a silyloxybutadiene to a bromoquinone
as the key step.9
Department of Chemistry, University of Auckland, 23 Symonds Street,
Auckland, New Zealand. E-mail: m.brimble@auckland.ac.nz
† Electronic supplementary information (ESI) available: Full experimental
data, nOe data and spectra of compounds 2, 5, 7, 8, 10a and 10b. See DOI:
10.1039/b905077a
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2599–2603 | 2599
©

View Article Online
The synthesis of 2 was hence conducted as shown in Scheme 2.
Although the Hauser–Kraus annulation between cyanophthalide
10 and (-)-enone 46 initially proved troublesome, it was eventually
Undeterred by these disappointing results, it was decided at
this stage to attempt the proposed late stage homocoupling of an
advanced aromatic intermediate instead of the pyranonaphtho-
quinone 2. Naphthopyran 7 was identified as the best candidate.
Thus, upon subjecting the naphthopyran 7 to a similar array
of oxidative coupling conditions, the one electron oxidant cerium
ammonium nitrate (CAN) proved the most promising.14 During
the CAN oxidation, complete consumption of the starting material
7 was only evident upon addition of three equivalents of the
oxidant, resulting in the formation of a very polar product.
Upon purification an orange solid was isolated that showed only
3
found that isolating the crude annulation product after a very short
reaction time (3–4 min) and immediately subjecting it to reductive
methylation conditions furnished naphthalene 5 in modest overall
yield. Treatment of 5 with excess TBAF for 4 days effected silyl
group removal with concomitant cyclisation and the resulting
crude lactol 6 was stereoselectively reduced with trifluoroacetic
acid and triethylsilane delivering the cis-1,3-dimethyl pyran 7 as a
single diastereomer in good overall yield.
The 1,3-stereochemistry was confirmed unequivocally by the
observed NOE correlation between the axial protons at C1 and
C3 on the pyran ring, a result expected due to the predicted
pseudoaxial delivery of hydride during the reduction step.11
Oxidative demethylation with phenyliodine bis(trifluoroacetate)12
in aqueous acetonitrile delivered ventiloquinone L methyl ether
(7-methoxyeleutherin) 8 for which the spectroscopic data were
identical to that reported for material prepared from a semi-
synthesis of 8 via oxidation of (+)-karwinaphthol B.13 Finally,
selective monodeprotection of 8 with boron trichloride delivered
(+)-ventiloquinone L 2 which was identical in all aspects to the
natural (+)-ventiloquinone L5 and synthetic (+)-ventiloquinone L9
obtained from the previous synthesis. Based on these results, we
further confirm the absolute stereochemistry present in 2 to be
1R, 3S.
1
one aromatic proton by H NMR, suggesting dimerisation had
occurred. The 13C spectrum showed the requisite amount of peaks
required for dimer 9, several of which had satellite peaks suggesting
the existence of diastereomers. Careful separation of the mixture
was carried out affording two distinct symmetrical atropisomers
10a and 10b in a 1 : 1 ratio suggesting that the chirality present
on the pyran ring in 7 had not exerted any chiral induction in the
biaryl forming step (Scheme 3).
X-Ray crystallographic analysis was conducted on pure 10a
which revealed the configuration at the C6–C6¢ biaryl bond to be
R but unfortunately the existence of conformers around the pyran
area of 10a precluded a complete crystallographic analysis. There-
fore, it should be noted that this study cannot be used to definitively
confirm the configuration of the biaryl bond. NOE studies on the
(S)-atropisomer 10b established the remaining aromatic proton
had a strong correlation with both methoxy groups on the aryl
ring, indicating that dimerisation had unfortunately not occurred
through C8 as proposed in our biomimetic hypothesis and as
is necessary for construction of the cardinalin 3 1 framework.
Rather, naphthopyran 7 had surprisingly dimerised exclusively at
Next, attention turned towards the key dimerisation step. Un-
fortunately, despite subjecting 2 to a plethora of known phenolic
oxidative dimerisation conditions including basic ferricyanide,
FeCl3, PbIV salts, CuI/O2, hn/O2 and cerium ammonium nitrate
(CAN), no cardinalin 3 1 was ever isolated. Similarly, attempted
dimerisation of 8 failed to deliver dimer 9 (Scheme 2).
1
C6. Also, the chemical shifts for the aromatic proton in the H
Scheme 2 Synthesis of (+)-ventiloquinone L 2 and failed dimerisation.
2600 | Org. Biomol. Chem., 2009, 7, 2599–2603 This journal is The Royal Society of Chemistry 2009
©

View Article Online
cardinalin 3 1. Thus, the 1 : 1 mixture of atropisomers 10a and
10b was treated with boron trichloride affording the atropisomers
12a and 12b also in a 1 : 1 ratio in good yield. The synthesis of
12a and 12b constitutes the total synthesis of the C6 regioisomer
of cardinalin 3 1 (Scheme 5).
In conclusion, an efficient synthesis of (+)-ventiloquinone L 2
has been achieved with a longest linear sequence of 7 steps further
confirming the 1R,3S stereochemistry of the natural product.
Although attempts to dimerise 2 directly were unsuccessful,
mild oxidation of naphthopyran 7 triggered homodimerisation
at C6 furnishing 10a and 10b. To the best of our knowledge,
this high yielding oxidative dimerisation of a naphthopyran or
pyranonaphthoquinone at C6 is extremely rare and the only
other similar reactions are the isolation of traces (2%) of a
C6 dimer during a semi-synthesis of actinorhodin from its
respective monomer15 and the borate promoted dimerisation of
quinone A.16 The mild biaryl formation demonstrated herein
provides interesting insight into the possible biosynthesis of
C6-symmetrical dimeric naphthopyran natural products such
as xylindein17 or viridotoxin18 from their respective monomers.
Furthermore, selective deprotection of 10a and 10b delivered 12a
and 12b, the C6 regioisomer of cardinalin 3 1. Our failure to
successfully dimerise pyranonaphthoquinones 2 or 8 combined
with the high yielding dimerisation of naphthopyran 7 exclusively
at C6 somewhat contradicted our proposal that cardinalin 3 1
is constructed in Nature from a late stage homocoupling of
ventiloquinone L 2 or a closely related compound. Also, due
to the facile and mild nature of the successful dimerisation of
naphthopyran 7 at C6, we propose that pyranonaphthoquinone
dimers of type 10a, 10b, 12a and 12b may well be as yet
undiscovered natural products.
Scheme 3 Homodimerisation of naphthopyran 7.
Experimental section
NMR of the aR and aS atropisomers 10a and 10b were observed
at d = 6.757 and 6.764 ppm respectively, significantly further
upfield than the aromatic protons observed at d = 7.53 ppm in
( )-cardinalin dimethyl ether 9 synthesised during de Koning’s
racemic synthesis of 1.4 The 13C chemical shifts of the remaining
aromatic methine carbons in 10a, 10b and 9 also showed significant
differences (Scheme 3).
Next, we set out to define the order of events in this in-
teresting dimerisation. Thus, in order to gauge whether biaryl
bond formation was occurring before or after the oxidative
demethylation, naphthopyran 7 was treated with 1 equivalent
of CAN. Surprisingly, only dimers 10a and 10b were isolated,
again as a 1 : 1 mixture along with substantial amounts of
starting material 7. None of the dimeric naphthopyran 11 was
ever observed. However, none of the dimers 10a or 10b were
isolated when treating ventiloquinone methyl ether 8 with varying
amounts of CAN, indicating that the oxidation of 7 to 8 is
not the first step in the CAN mediated reaction of 7 to 10a
and 10b. These results suggest that in the transformation of
naphthopyran 7 to 10a and 10b, the biaryl bond is formed in the
first step followed by immediate oxidation of the resulting electron
rich dimeric naphthopyran 11 affording the dimers 10a and 10b
(Scheme 4).
(1R,1¢R,3S,3¢S,6R)-7,7¢,9,9¢-Tetramethoxy-1,1¢,3,3¢-tetramethyl-
3,3¢,4,4¢-tetrahydro-1H,1¢H-6,6¢-bibenzo[g]isochromene-
5,5¢,10,10¢-tetraone (10a) and (1R,1¢R,3S,3¢S,6S)-7,7¢,9,
9¢-tetramethoxy-1,1¢,3,3¢-tetramethyl-3,3¢,4,4¢-tetrahydro-1H,
1¢H-6,6¢-bibenzo[g]isochromene-5,5¢,10,10¢-tetraone (10b)
Naphthopyran 7 (48 mg, 0.14 mmol) was taken up in acetonitrile
(4 mL) and a solution of cerium(IV) ammonium nitrate (245 mg,
0.45 mmol) in distilled water (2 mL) was added. The reaction mix-
ture was stirred for 30 min at rt and water (12 mL) was then added.
The aqueous layer was extracted with ethyl acetate (3 ¥ 30 mL)
and the combined organic layers dried over anhydrous magnesium
sulfate, filtered and concentrated in vacuo. The resulting residue
was purified by flash chromatography eluting with hexanes–ethyl
acetate (1 : 3 then 100% ethyl acetate) to give the title compounds
as an orange solid and a 1 : 1 mixture of atropisomers; (30.7 mg,
0.051 mmol, 71%). Purification of this mixture by further flash
chromatography eluting with hexanes–ethyl acetate (1 : 1 then
2 : 3) gave pure (R)-atropisomer 10a (Rf 0.42, ethyl acetate) as a
yellow solid (14 mg, 0.023 mmol, 34%) and pure (S)-atropisomer
10b (Rf 0.35, ethyl acetate) as a yellow solid (15 mg, 0.025 mmol,
36%).
◦
10a Mp 270–272 C; [a]D24 +852.9 (c 0.13, CH2Cl2); nmax (neat)
At this stage it was decided to selectively deprotect 10a and
10b in order to complete the total synthesis of a regioisomer of
2928, 2851, 1737, 1649, 1634, 1582, 1551, 1456, 1433, 1341, 1300,
1253, 1211, 1149 cm-1; dH (300 MHz, CDCl3) 1.27 (6 H, d, J 6.3,
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2599–2603 | 2601
©

View Article Online
Scheme 4 Route for transformation of 7 to 10a and 10b.
(2 ¥ CH), 114.8 (2 ¥ C), 120.9 (2 ¥ C), 132.3 (2 ¥ C), 139.8 (2 ¥ C),
=
148.0 (2 ¥ C), 161.3 (2 ¥ C), 161.6 (2 ¥ C), 182.7 (2 ¥ C O), 184.8
+
=
(2 ¥ C O); m/z (EI+) 602 (100%, M ), 587 (10), 260 (10), 43 (80);
HRMS (EI+, M+) found 602.2150, calc for C34H34O10 602.2152.
◦
10b Mp 127–129 C; [a]D24 +108.1 (c 0.16, CH2Cl2); nmax (neat)
2928, 2851, 1737, 1649, 1634, 1582, 1551, 1456, 1433, 1341, 1300,
1253, 1211, 1149 cm-1; dH (300 MHz, CDCl3) 1.26 (6 H, d, J 6.0,
2 ¥ CH2CHMe), 1.53 (6 H, d, J 6.6, 2 ¥ Me), 1.94 (2 H, ddd,
J 3.8, 10.3 and 18.1, 2 ¥ CHaxH), 2.50 (2 H, t, J 2.6 and 18.1,
2 ¥ CHHeq), 3.49 (2 H, m, 2 ¥ H-3), 3.73 (6 H, s, 2 ¥ OMe), 4.05
(6 H, s, 2 ¥ OMe), 4.82 (2 H, m, 2 ¥ H-1), 6.764 (2 H, s, 2 ¥ Ar–H);
dC (75 MHz, CDCl3) 20.6 (2 ¥ Me), 21.2 (2 ¥ Me), 29.8 (2 ¥ CH2),
56.1 (2 ¥ OMe), 56.3 (2 ¥ OMe), 68.8 (2 ¥ CH), 70.3 (2 ¥ CH), 99.4
(2 ¥ CH), 114.9 (2 ¥ C), 121.4 (2 ¥ C), 131.3 (2 ¥ C), 139.6 (2 ¥ C),
=
148.2 (2 ¥ C), 161.3 (2 ¥C), 162.6 (2 ¥ C), 183.1 (2 ¥ C O), 184.8
+
=
(2 ¥ C O); m/z (EI+) 602 (100%, M ), 587 (10), 260 (10), 43 (80);
HRMS (EI+, M+) found 602.2150, calc for C34H34O10 602.2152.
Acknowledgements
We thank the Royal Society of New Zealand Marsden fund for
financial support.
References
Scheme 5 Synthesis of the C6 regioisomer of cardinalin 3 1.
1 For a review regarding isolation, structure and synthesis of pyranon-
aphthoquinone antibiotics see: J. Sperry, P. Bachu and M. A. Brimble,
Nat. Prod. Rep., 2008, 25, 376.
2 (a) M. Gill, M. S. Buchanan and J. Yu, J. Chem. Soc., Perkin Trans. 1,
1997, 919; (b) M. Gill, M. S. Buchanan and J. Yu, Aust. J. Chem., 1997,
50, 1081.
3 (a) M. A. Brimble, J. S. Gibson, J. J. P. Sejberg and J. Sperry, Synlett,
2008, 867; (b) J. Sperry, J. S. Gibson, J. J. P. Sejberg and M. A. Brimble,
Org. Biomol. Chem., 2008, 6, 4261.
2 ¥ CH2CHMe), 1.56 (6 H, d, J 6.6, 2 ¥ Me), 1.97 (2 H, ddd,
J 3.8, 10.2 and 18.2, 2 ¥ CHaxH), 2.50 (2 H, dt, J 2.6 and 18.2, 2 ¥
CHHeq), 3.48 (2 H, m, 2 ¥ H-3), 3.74 (6 H, s, 2 ¥ OMe), 4.05 (6
H, s, 2 ¥ OMe), 4.81 (2 H, m, 2 ¥ H-1), 6.757 (2 H, s, 2 ¥ Ar–H);
dC (75 MHz, CDCl3) 20.7 (2 ¥ Me), 21.2 (2 ¥ Me), 29.9 (2 ¥ CH2),
56.1 (2 ¥ OMe), 56.3 (2 ¥ OMe), 68.8 (2 ¥ CH), 70.3 (2 ¥ CH), 99.6
2602 | Org. Biomol. Chem., 2009, 7, 2599–2603
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
4 S. Govender, E. M. Mmutlane, W. A. L. van Otterlo and C. B. de
Koning, Org. Biomol. Chem., 2007, 5, 2433.
5 S. R. Jammula, S. B. Pepalla, H. Telikepalli, K. V. J. Rao and R. H.
Thompson, Phytochemistry, 1991, 30, 3741.
6 J. S. Gibson, O. Andrey and M. A. Brimble, Synthesis, 2007,
2611.
11 G. A. Kraus, M. T. Molina and J. A. Walling, J. Chem. Soc., Chem.
Commun., 1986, 1568.
12 H. Tohma, H. Moriaka, Y. Harayama, M. Hashizume and Y. Kita,
Tetrahedron Lett., 2001, 42, 6899.
13 L. A. Mitscher, S. R. Gollapudi, D. S. Oburn and S. Drake, Phytochem-
istry, 1985, 24, 1681.
7 (a) F. M. Hauser and R. P. Rhee, J. Org. Chem., 1978, 43, 178; (b) G. A.
Kraus and H. Sugimoto, Tetrahedron Lett., 1978, 19, 2263; (c) For
recent reviews on the Hauser–Kraus annulation see: K. Rathwell and
M. A. Brimble, Synthesis, 2007, 643; (d) D. Mal and P. Pahari, Chem.
Rev., 2007, 107, 1892.
8 E. M. Mmutlane, J. P. Michael, I. R. Green and C. B. de Koning, Org.
Biomol. Chem., 2004, 2, 2461.
14 The use of CAN to effect biaryl bond formation is relatively rare,
see: (a) A. Dhakshinamoorthy, Synlett, 2005, 3014; (b) V. Nair, L.
Balagopal, R. Rajan and J. Mathew, Acc. Chem. Res., 2004, 37, 21 and
references therein.
15 H. Laatsch, Liebigs Ann. Chem., 1987, 297.
16 D. W. Cameron, H. W.-S. Chan and E. M. Hildyard, J. Chem. Soc. C,
1966, 1836.
9 L. M. Tewierik, C. Dimitriadis, C. D. Donner, M. Gill and B. Willems,
Org. Biomol. Chem., 2006, 4, 3311.
17 Y. Saikawa, T. Watanabe, K. Hashimoto and M. Nakata, Phytochem-
istry, 2000, 55, 237.
10 M. A. Brimble, C. L. Flowers, J. K. Hutchinson, J. E. Robinson and
M. Sidford, Tetrahedron, 2005, 61, 10036.
18 K. Suzuki, K. Nozawa, S. Nakajima and K. Kawai, Chem. Pharm.
Bull., 1990, 38, 3180.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2599–2603 | 2603
©