Biomimetic synthesis of dimeric metabolite acremine g via a highly regioselective and stereoselective Diels-Alder reaction

Article abstract of DOI:10.1021/ol901004e

The dimeric metabolite acremine G was synthesized featuring a highly regioselective and stereoselective Diels-Alder reaction between a TBS-protected hydroquinone diene and a structurally related alkenyl quinone. The major endo [4 + 2] adduct slowly transforms to acremine G by the atmospheric air under the deprotection conditions (in situ generated HF).

Full text of DOI:10.1021/ol901004e

ORGANIC
LETTERS
2009
Vol. 11, No. 14
2988-2991
Biomimetic Synthesis of Dimeric
Metabolite Acremine G via a Highly
Regioselective and Stereoselective
Diels-Alder Reaction
Elias Arkoudis, Ioannis N. Lykakis, Charis Gryparis, and Manolis Stratakis*
Department of Chemistry, UniVersity of Crete, Voutes 71003 Iraklion, Greece
stratakis@chemistry.uoc.gr
Received May 8, 2009
ABSTRACT
The dimeric metabolite acremine G was synthesized featuring a highly regioselective and stereoselective Diels-Alder reaction between a
TBS-protected hydroquinone diene and a structurally related alkenyl quinone. The major endo [4 + 2] adduct slowly transforms to acremine
G by the atmospheric air under the deprotection conditions (in situ generated HF).
Acremines A-F,1 G,2 and H-N3 (Figure 1) are metabolites
isolated from the cultures of the fungus Acremonium bys-
soides. They inhibit the germination of sporangia of Plas-
mopara Viticola. Some of them exhibit modest cytotoxic
activity against the tumor cell line H460, while acremine A
inhibits lipoxygenase. Acremine A has also been isolated4
from the methanolic extract of Periploca aphylla. Its initially
misassigned structure was corrected later.5 Of particular
synthetic interest is acremine G which biosynthetically could
derive from a Diels-Alder reaction between the so far
nonisolated hydroquinone diene 1 and alkenyl quinone 2
(Scheme 1),6 accompanied by an intermolecular oxidative
coupling in the [4 + 2] adduct. Biomimetic Diels-Alder
reactions and dimerizations have been proposed for a plethora
of natural product biosyntheses and have been accomplished
aesthetically7 in several cases by synthetic organic chemists.
Monomer 2 which resembles acremines A and B could
form 1 via dehydration/oxidation. A similar biosynthetic
proposal had been postulated for the structurally related
allomicrophyllone.8 To test this postulated [4 + 2] biogenetic
(6) For a recent example of employing a Diels-Alder reaction between
a quinone and a diene, as a key step towards the biomimetic synthesis of
cordiaquinones in our lab, see: Arkoudis, E.; Stratakis, M. J. Org. Chem.
2008, 73, 4484–4490.
(7) Selected examples: (a) Nicolaou, K. C.; Snider, S. A.; Montagnon,
T.; Vassilikogiannakis, G. Angew. Chem., Int. Ed. 2002, 41, 1668–1698.
(b) Stocking, E. M.; Williams, R. M. Angew. Chem., Int. Ed. 2003, 42,
3078–3115. (c) Mehta, G.; Pan, S. C. Org. Lett. 2004, 6, 3985–3988. (d)
Zhang, W.; Luo, S.; Fang, F.; Chen, Q.; Hu, H.; Jia, X.; Zhai, H. J. Am.
Chem. Soc. 2005, 127, 18–19. (e) Gagnepain, J.; Castet, F.; Quideau, S.
Angew. Chem., Int. Ed. 2007, 46, 1533–1535. (f) Dong, S.; Zhu, J.; Porco,
J. A., Jr. J. Am. Chem. Soc. 2008, 130, 2738–2739. (g) Layton, M. E.;
Morales, C. A.; Shair, M. D. J. Am. Chem. Soc. 2002, 124, 773–775. (h)
Porco, J. A., Jr.; Su, S.; Bardhan, S.; Rychnovsky, S. D. Angew. Chem.,
Int. Ed. 2006, 45, 5790–5792. (i) Vanderwal, C. D.; Vosburg, D. A.; Weiler,
S.; Sorensen, E. J. J. Am. Chem. Soc. 2003, 125, 5393–5407. (j) Nicolaou,
K. C.; Lim, Y. H.; Becker, J. Angew. Chem., Int. Ed. 2009, 48, 3444–
3448.
(1) Assante, G.; Dallavalle, S.; Malpezzi, L.; Nasini, G.; Burruano, S.;
Torta, L. Tetrahedron 2005, 61, 7686–7692.
(2) Arnone, A.; Nasini, G.; Panzeri, W.; de Pava, O. V.; Malpezzi, L.
J. Nat. Prod. 2008, 71, 146–149.
(3) Arnone, A.; Assante, G.; Bava, A.; Dallavalle, S.; Nasini, G.
Tetrahedron 2009, 65, 786–791.
(4) Aziz-ur, R.; Malik, A.; Riaz, N.; Nawaz, H. R.; Ahmad, H.; Nawaz,
S. A.; Choudhary, M. I. J. Nat. Prod. 2004, 67, 1450–1454.
(5) Hoarau, C.; Pettus, T. R. R. Org. Lett. 2006, 8, 2843–2846.
(8) Yamamura, S.; Simpol, L. R.; Ozawa, K.; Ohtani, K.; Otsuka, H.;
Kasai, R.; Yamasaki, K.; Padolina, W. G. Phytochemistry 1995, 39, 105–
110.
10.1021/ol901004e CCC: $40.75
Published on Web 06/18/2009
 2009 American Chemical Society

Scheme 2. Attempted Synthesis of Diene Monomer 1
Figure 1. Acremines A-N.
Diene 1 is extremely labile and decomposes within a few
minutes. In one of our numerous attempts, luckily enough 1
was seen by H NMR before undergoing decomposition. It
1
pathway, we undertook the synthesis of the proposed
monomers 1 and 2. Regarding the attempted synthesis of
diene 1 (Scheme 2), toluhydroquinone was used as the
starting material. The TBS-protected toluhydroquinone 3 was
iodinated selectively at C-5 with I2/CF3COOAg to form 4
in 85% yield, while the aryl iodide 4 underwent a Heck
coupling (Pd(OAc)2, (o-tolyl)3P, Et3N, DMF) with 2-meth-
ylbut-3-en-2-ol9 to afford 5 in 67% isolated yield. The allylic
alcohol 5 was easily dehydrated to diene 6 on treatment with
acetyl chloride/pyridine.9 We observed that even on standing
5 slowly dehydrates to 6. While the silyl-protected diene 6
is quite stable for a long time, all initial attempts to deprotect
the TBS groups (TBAF, Pd(CH3CN)2Cl2,10 catal. TMSBr,11
LiOAc,12 CuBr2,13 and other) failed to provide any hydro-
quinone diene. The desired diene 1 is, as expected, sensitive
to acidic or basic conditions and rather polymerizes very fast
leaving behind a tary residue. By using a mild desilylation
procedure modified by Hudson14 (KF, HBr/CH3COOH, dry
DMF), the deprotection proceeded efficiently after 4-5 h at
room temperature.
partially transforms, among other unidentified products and
tar, to chromene 7, whose spectral data are in agreement
with those of acremine C1 (Figure 1). The cyclization of 1
to 7 is obviously catalyzed by acidic impurities. Hence, we
decided to use the silyl-protected diene 6 en route to
accomplish the [4 + 2] scenario shown in Scheme 1.
The synthesis of the desired dienophile 2 was accom-
plished in five steps (Scheme 3) from the commercially
Scheme 3. Synthesis of the Alkenyl Quinone 2
Scheme 1
. Postulated Biosynthesis of Acremine G via a
Tandem Diels-Alder Dimerization/Oxidation
Org. Lett., Vol. 11, No. 14, 2009
2989

available 2-methyl-1,4-dimethoxybenzene. Thus, prenyla-
tion15 at C-5 under Friedel-Crafts conditions (prenol, BF3
etherated) afforded 8,16 which after dihydroxylation (cat.
OsO4, NMO), protection of the 1,2-diol 9 as carbonate 10
with 1,1′-carbonyldiimidazole (CDI), and reaction of 10 with
ceric ammonium nitrate, afforded quinone 11 in 52% yield
over the three last steps. Finally, treatment of 11 with Hunig’s
base17 (15 equiv) in refluxing benzene cleanly afforded 2 in
84% isolated yield. It should be noted that regarding the last
step, the reactant dilution must be kept as low as 0.01 M,
otherwise a mixture of products is formed in low yield.
Having in our hands diene 6 and dienophile 2, we
performed their [4 + 2] cycloaddition in refluxing toluene.
The reaction was complete after 24 h to afford a mixture of
12 (endo product) and 13 (exo product) in 78% yield and in
a relative ratio 12/13 ∼ 5/1. The structural assignment of
the products was performed by NOE experiments, with some
key enhancements shown in Scheme 4. The selective
1 possibly dictates the regiochemical and stereochemical
outcome. On the other hand, the regiochemical preference
for cycloaddition on the vinyl-substituted double bond of
2-monovinyl-substituted 1,4-quinones is known,18 as a result
of the larger coefficients of its LUMO (semiempirical
calculations).
The endo adduct 12, which from the stereochemical point
of view is the suitable precursor of acremine G, underwent
deprotection (6.0 equiv of KF, 2.5 equiv of HBr 33% in
CH3COOH, dry DMF).14 These deprotection conditions (in
situ generated HF) are the only effective ones, as all
variations discussed above during the attempted deprotection
of 6 failed, providing either a mixture of products and/or
tar. After 1 h, a new product had been primarily formed
(TLC), which was isolated and characterized by NMR as
the selectively monodeprotected on the less hindered -OTBS
group 14 (NOE interaction of H-6 with the TBS group,
Scheme 5). The cis stereochemistry among the fused six-
Scheme 5
. Deprotection of the Diels-Alder Adduct 12 with in
Scheme 4. Diels-Alder Reaction between 6 and 2 and the
Stereochemical Assignment of the Cycloadductsa
Situ Generated HF, Leading to Acremine G
membered rings was retained in 14 as revealed by the NOE
enhancement (4.4%) among the hydrogen atoms on C-3′ and
the olefinic hydrogens on the alkenyl (C7′-C8′) side chain.
Leaving the reaction mixture for a longer time, a new more
polar product gradually appeared. After 24 h, it was isolated
and ascribed as the dideprotected starting material. The
completely deprotected 12 was a mixture of two epimers in
a ratio ∼5/1 with the major assigned as the trans-fused 15
(Scheme 5). In 15, the lack of NOE enhancement between
the hydrogen atom H-3′ and the alkenyl (C7′-C8′) side chain
is characteristic, while a strong enhancement of the hydrogen
atom on C-3 was observed (13.7%) upon irradiating the
hydrogen atom on C-3′, in agreement with our stereochemical
a The atom numbering of 12 and 13 was arbitrarily considered after the
numbering of acremine G in the isolation paper.
formation of 12 among seven other possible regio/stereoi-
somers could be explained considering the endo transition
state (Scheme 4), where a nonbonding interaction between
the alkenyl chain of quinone and the aryl substituent of diene
(9) Harrington, P. J.; Hegedus, L. S.; McDaniel, K. F. J. Am. Chem.
Soc. 1987, 109, 4335–4338.
(10) Wilson, N. S.; Keay, B. A. Tetrahedron Lett. 1996, 37, 153–156.
(11) Shah, S. T. A.; Guiry, P. J. Org. Biomol. Chem. 2008, 6, 2168–
2172.
2990
Org. Lett., Vol. 11, No. 14, 2009

assignment. Obviously, the cis stereochemistry among the
fused rings observed in 12 or 14 was not retained in the
fully deprotected cycloadduct 15 due to an epimerization
under the slightly acidic reaction conditions. The minor
dideprotected product 16 could not be separated from 15 by
column chromatography, and we tentatively assigned its
structure as the epimer of 15 on C-3′. However, in the mean
time, besides monoprotected 14 and fully deprotected 15/
16, a new product with an intermediate polarity among them
started to form and became the major after 24 h, while after
36 h it was almost the only reaction product. We were
pleased to realize that the newly formed product was
acremine G19 (71% isolated yield from 12). Additionally,
acremine G was also exclusively formed after 24 h on leaving
the mixture of 15/16 in DMF, having added acetic acid (5
equiv).
Scheme 6. Postulated Mechanism for the Oxidation of 15/16 to
Acremine G by Electron Transfer to Molecular Oxygen
We postulate that the oxidation of 15/16 to acremine G
occurs by molecular oxygen via a radical pathway (Scheme
6). A reasonable mechanism requires formation of the radical
cation of the hydroquinone moiety in 15/16 by electron
transfer either intramolecularly from the semiquinone moiety
or intermolecularly by molecular oxygen (affording super-
oxide anion). The radical cation eliminates a proton, and the
resulting oxygen centered radical couples with the double
bond of the common enol form of 15/16. Finally, the cyclized
C-centered radical reacts with the superoxide anion to form
a labile hydroperoxy hemiacetal, which eliminates H2O2 to
yield acremine G. We postulate that the slightly acidic
conditions during the deprotection process facilitate the
oxidative coupling through catalyzing the formation of the
necessary enol. Notably, we did not observe any product
resulting from the coupling among C3 and C3′ that would
provide a compound related to hydroxymicrophyllone.8 A
profound reason is that a more strained system is thus
generated, relative to the six-membered ring coupling which
generates acremine G.
In conclusion, we have presented an efficient biomimetic
route to acremine G featuring a highly regioselective and
stereoselective Diels-Alder reaction. Synthetic studies to-
ward microphyllones8 (structurally related to acremine G
natural products) as well as an exploration of the hydro-
quinone-ketone oxidative coupling are in progress.
(12) Wang, B.; Sun, H.-X.; Sun, Z.-H. J. J. Org. Chem. 2009, 74, 1781–
1784.
Acknowledgment. E.A. thanks IKY for a 4 year fellow-
ship. We thank professor A. Berkessel (Cologne University)
for obtaining some HRMS.
(13) Bhatt, S.; Nayak, S. K. Tetrahedron Lett. 2006, 47, 8395–8399.
(14) Suleman, A. S.; Tillekeratne, L. M. V.; Hudson, R. A. Synth.
Commun. 2001, 31, 3303–3308.
(15) Misiti, D.; Moore, H. W.; Folkers, K. Biochemistry 1965, 4, 1156–
1160.
Supporting Information Available: Experimental details
for the synthesis of monomers 2 and 6 and their cycload-
(16) Scheepers, B. A.; Klein, R.; Davies-Coleman, M. T. Tetrahedron
Lett. 2006, 47, 8243–8246.
1
dition. Copies of H, 13C NMR, HRMS, and MS spectra of
(17) Lindsey, C. C.; Gomez-Diaz, C.; Villalba, J. M.; Pettus, T. R. R.
Tetrahedron 2002, 58, 4559–4565.
key compounds and reactions. This material is available free
of charge via the Internet at http://pubs.acs.org.
(18) Irngartinger, H.; Stadler, B. Eur. J. Org. Chem. 1998, 605–626.
(19) For a comparison of the NMR spectroscopic data of natural and
synthetic acremine G, see the Supporting Information.
OL901004E
Org. Lett., Vol. 11, No. 14, 2009
2991