5560
G. Appendino et al. / Tetrahedron Letters 50 (2009) 5559–5561
To combine proof of principle and biological relevance, we have
complex mechanistic scenario. Thus, when mixtures of adducts
were obtained, only formation of bis-indolyl derivatives was ob-
served, suggesting that indole can compete with the b-dicarbonyl
nucleophile for attack to the aldehyde, or, alternatively, that in-
dole can displace b-dicarbonyls from their heterodimeric ad-
ducts.8 Furthermore, formation of homodimeric indolyl adducts
was more marked with aromatic than with aliphatic aldehydes,
and with triacetic acid lactone than with 4-hydroxycoumarin.
To improve the yield of heterodimeric adducts, various solvent
systems and catalysts were screened, eventually discovering that
the use of a biphasic reaction system (chloroform–water 1:1)
could steer the reaction toward the formation of heterodimeric
adducts, solving, at least from a preparative standpoint, this issue
(Table 1). Thus, while the reaction of p-nitrobenzaldehyde with 4-
hydroxycoumarin and indole afforded an equimolecular mixture
of the heterodimeric adduct 8d and its corresponding bis-indolyl
adduct, switching to CHCl3–water provided the heterodimeric ad-
duct 8d in 85% yield. Occasionally, the heterodimeric adducts di-
rectly precipitated from the reaction mixture, while gravity
column chromatography was required when a homogeneous
reaction mixture was obtained.9
taken inspiration from the structures of 1–3, selecting 4-hydroxy-
coumarin (5) triacetic acid lactone (6), and indole (7) as the react-
ing partners. Aldehydes representative of various electronic and
steric conditions were employed (acetaldehyde, heptanal, pivala-
dehyde, benzaldehyde and its p-methoxy- and p-nitro derivatives),
while chloroform was used to optimize the reaction conditions be-
cause of the possibility to benefit from real-time NMR control
when the deuterated version of the solvent was employed in
exploratory experiments.
Mechanistic considerations point to the possibility of obtaining
heterodimeric adducts from the reaction of aldehydes, b-dicarbon-
yls, and arenes. Thus, the hard aldehyde carbonyl is expected to re-
act preferentially with the more localized nucleophilic double
bond of an enol (Scheme 1, A) rather than with the softer and more
dispersed
p-system of an electron-rich aromatic. Conversely, the
resulting Knoevenagel adduct (Scheme 1, B) is a softer electrophile,
and should therefore react with an electron-rich aromatic nucleo-
phile better than with an enolized b-dicarbonyl.7
Preliminary experiments (Table 1), while largely confirming
these insights, also afforded several clues indicative of a more
Ultimately, heterodimeric adducts could be obtained from all
aldehydes investigated except from pivaladehyde. This behavior
is presumably due to steric hindrance from the bulky t-butyl
group, that prevents the addition, in a Michael fashion, of a second
nucleophilic species.10
The role of water to steer the reaction towards the formation of
the heterodimeric adducts is difficult to rationalize, but goes be-
yond that of a mere precipitating agent for the generally poorly sol-
uble heterodimeric adducts. Comparison of the results observed
with heptanal and acetaldehyde upon reaction with indole (7)
and triacetic acid lactone (6) exemplifies this issue. Thus, while
acetaldehyde afforded a copious precipitate (75% yield) of the hete-
rodimeric adduct 9a when the reaction of heptanal was carried out
O
O
R
O
O
O
H
R
-H2O
O
H
N
H
O
7
B
A
R
OH
O
O
N
H
under identical conditions,
a homogeneous reaction mixture
C
mainly containing heptyliden-bis-indolylmethane was obtained.
However, when the reaction was performed in 1:1 chloroform–
water, heptanal afforded a heterodimeric adduct (9b) as the major
reaction product (47% isolated yield).
Scheme 1. Mechanism of the multicomponent reaction of aldehydes, b-dicarbon-
yls, and indole.
Taken together, our results, while pointing to new applications
of multicomponent reactions and providing a straightforward en-
try into an interesting class of bioactive compounds, also raise a
series of mechanistic issues worth further investigation.
Table 1
Heterodimeric adducts from the multicomponent reaction of aldehydes, indole and 4-
hydroxycoumarin (compounds 8a–e) or triacetic acid lactone (compounds 9a–e)
R
OH
Acknowledgments
O
O
N
H
We thank MIUR (Fondi ex-40%) for financial support. We are
grateful to Mrs. Anna Maria Bottani for her help throughout this
study.
R
Conditions (time, yield)*
8a
8b
8c
8d
8e
CH3
C6H13
C6H5
pNO2C6H4
pOMeC6H4
a (5 h, 88%)
a (48 h, 70%)
a (24 h, 41%) b (48 h, 57%)
a (24 h, 25%) b (48 h, 85%)
a (24 h, 18%) b (48 h, 28%)
References and notes
1. Stahmann, M. A.; Huebner, C. F.; Link, K. P. J. Biol. Chem. 1941, 138, 513–527.
2. For a review, see: Safe, S.; Papineni, S.; Chintharlapalli, S. Cancer Lett. 2008, 269,
326–338.
3. Watsuji, T. O.; Yamada, S.; Yamabe, T.; Watanabe, Y.; Kato, T.; Saito, T.; Ueda,
K.; Beppu, T. Appl. Environ. Microbiol. 2007, 73, 6159–6165.
R
OH
4. Dicoumarol was synthesized by Hanschutz as early as in 1909 (Hanschutz, R.
Liebigs Ann. Chem. 1909, 367, 169–218).
5. Some b-dicarbonyl-heteroaryl cross dimers have been prepared in a two-step
protocol from the condensation of heteroaryl aldehydes and b-dicarbonyls,
followed by hydrogenation of the resulting Knoevenagel adducts with Pd/
CaCO3 (Velezheva, V. S.; Erofeev, Y. V.; Suvorov, N. N. Zh. Org. Khim. 1973, 9,
185–9).
6. (a) Appendino, G.; Ottino, M.; Marquez, N.; Bianchi, F.; Ballero, M.; Sterner, O.;
Fiebich, B. L.; Muñoz, E. J. Nat. Prod. 2007, 70, 608–612; (b) Rosa, A.; Deiana, M.;
Atzeri, A.; Corona, G.; Melis, M. P.; Appendino, G.; Dessì, M. A. Chem. Biol.
Interact. 2007, 165, 117–126.
O
O
N
H
9a
9b
9c
9d
9e
CH3
C6H13
C6H5
a (6 h, 75%)
a (24 h, traces) b (24 h, 47%)
a (24 h, 21%) b (72 h, 26%)
a (24 h, traces) b (72 h, 59%)
a (24 h, 18%) b (72 h, 12%)
pNO2C6H4
pOMeC6H4
7. Compared to enols, electron-rich aromatics are more polarizable, easier to
oxidize, and therefore ‘softer’ (Ho, T. L. Chem. Rev. 1975, 75, 1–20).
*
a: CHCl3, 40 °C; b: CHCl3–water 1:1, 40 °C.