D. Trauner et al.
ular nucleophilic attack of a ter-
tiary amine onto an ortho-qui-
none, tautomerization, and fi-
nally oxa-6p-electrocyclization,
as shown in Scheme 4. The use
of silver(II) oxide proved essen-
tial to promote this reaction
cascade, as subjection of crude
15 to an ambient oxygen atmos-
phere along with other oxida-
tion conditions failed to oxidize
the catechol moiety and result-
ed in the recovery of pure 15.
Our efficient synthesis of exi-
guamine A yielded enough ma-
terial for extensive biological
testing and allowed us to deter-
mine the Ki of the natural prod-
uct. It also gave us an opportu-
nity to further explore the fasci-
nating chemistry of catechola-
mines. During attempts to im-
prove the yield of the final step,
we treated crude 15 with
a twenty-fold excess of silver-
(II) oxide. Under these condi-
Scheme 13. Proposed biosynthetic relationship of exiguamine A and B.
tions, we noticed the formation
of a new product in addition to
trace amounts of exiguamine A.
This compound was isolated in 45% yield and was identified
as exiguamine B, a hydroxylated derivative of exiguamine A
(Scheme 12). Notably, exiguamine B was not found when
exiguamine A was resubjected to our standard oxidative
conditions. Exiguamanie B contains structural features of
other important catecholamines, such as adrenaline and nor-
adrenaline, which are formed biosynthetically through enzy-
matic hydoxylations of the benzylic position.
At this stage, we became aware that the name of exigua-
mine A implies the existence of other exiguamines (exigua-
mine B, C, etc.). We therefore contacted the isolation group
with our spectral data and were happy to learn that our hy-
droxylated byproduct was indeed identical to a natural prod-
uct, namely exiguamine B, which had only been obtained in
minute quantities from natural sources. Using our synthetic
product, the relative stereochemistry of exiguamine B could
be clarified by NOE measurements.[9]
lowed by oxa-6p electrocyclization would then yield exigua-
mine B, in analogy to exiguamine A. In line with our hy-
pothesis, resubjection of exiguamine A to excess AgO did
not yield exiguamine B. This seems to rule out an alterna-
tive mechanism that would involve nucleophilic attack of
water onto a para-quinone methide.
We have previously demonstrated that tautomerizations
and electrocyclizations of the type shown in Scheme 13 are
surprisingly facile and are associated with low kinetic barri-
ers.[10] In order to probe the reversibility of the individual
steps proposed herein, we performed a density functional
theory study (B3LYP/6-31G**,[33–35] Jaguar 6.5).[36] The dif-
ferences in electronic energies (E) of 5, 6, 46, 47, and 48
were found to be no more than 12 kcalmolꢁ1 (see Support-
ing Information). By contrast, exiguamine A is energetically
favored over 6 by 30 kcalmolꢁ1. Once formed, 5 will, thus,
provide a complex mixture of tautomers 5, 6, 46, and 48,
which ultimately converges on exiguamine A. Since the oxa-
6p electrocyclization to afford dihydropyran 47 is reversible,
the pathway to exiguamine B is revealed only upon oxida-
tion to 49.
Proposed biosynthesis of exiguamine B: To account for the
formation of exiguamine B, both in the laboratory and in
nature, we propose that it is also a product derived from bis-
quinone 5, which is in equilibrium with its tautomers 6, 46,
and 48 (Scheme 13). A reversible oxa-6p electrocyclization
of 46 would then afford 47 and place an oxygen functionali-
ty at the benzylic position. In the presence of a large excess
of oxidant (20 instead of 10 equivalents of AgO), 47 is inter-
cepted via an oxidation to yield hydroxylated bisquinone 49.
Tautomerization to vinyl ortho-quinone methide 50, fol-
Conclusion
In summary, we have presented the evolution of a synthetic
plan that yielded two complex natural products with many
novel structural and functional features. Our synthetic
5004
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4999 – 5005