C. E. Gutteridge et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5682–5686
5683
all-carbon naphthalene or a nitrogen-containing quino-
line were prepared, since it was not clear from previous
reports which contributed more to antimalarial activi-
ty.4 A potential advantage of heterocyclic compounds
is reduced lipophilicity, which should improve aqueous
solubility—important since the modest water solubility
of the chalcone scaffold may be an obstacle to future
drug development.5 Substitution on the 1-phenyl ring
is known to significantly influence activity—dimethoxy
and especially dichloro substitution patterns are well
represented in the more active analogs synthesized to
date.4,6,7 This study therefore focused on such struc-
tures, including novel analogs containing methoxy and
halo-substituents in the same compound. Analogs
substituted at the 4-position of the 1-phenyl ring were
of particular interest since this site appears to be espe-
cially prone to metabolic oxidation.8
previously.4,6,7 The initial series of compounds prepared
(1–33) contain either a carbocyclic naphthalene or a het-
erocyclic quinoline as the R2-substituent, with either
one, two or three methoxy groups as the R1-substitu-
ent(s). Most of the naphthalene-derivatives were moder-
ately active (IC50 in the range 1–10 lM). In all cases
1-naphthyl compounds appear to be superior to their
corresponding 2-naphthyl isomers (comparison of com-
pounds 1 and 2, 6 and 7, 11 and 12, 15 and 16, 19 and
20, 23 and 24, 26 and 27, and 31 and 32). In these naph-
thalene series the various di- and tri-methoxy com-
pounds tested had similar activities, and were more
active than the corresponding mono-methoxy analogs.
As noted above, it was not clear from previous studies
whether the presence of a nitrogen in the R2-substituent
always enhanced activity.4 In this study, the nitrogen-con-
taining 3-quinolines appear more active than their corre-
sponding carbocyclic analogs, the 2-naphthalenes
(comparison of 4 and 2, 9 and 7, 13 and 12, 17 and 16,
21 and 20, 25 and 24, and 29 and 27), with just one excep-
tion (33 and 32). However, this was not the case with 4-
quinolines and their corresponding carbocyclic analogs,
the 1-naphthalenes (10 and 6, and 22 and 19 favor the
quinoline, whereas 14 and 11, 18 and 15, and 30 and 26
favor the naphthalene). Antimalarial activities of the 2-
quinoline-derivatives compare favorably with those of
their 3- and 4-quinoline, and 2-naphthalene analogs, a
finding that has not been described previously. In the 2-,
3-, and 4-quinoline series, dimethoxy compounds were
more active than the corresponding mono- and tri-meth-
oxy analogs. When the naphthalene was preferred over
the quinoline, that difference in activity appears no greater
than threefold. However, examples were found where the
activity of the quinoline was tenfold better than that of the
naphthalene. Based on these results a second series of
compounds, designed to explore halo and halomethoxy
R1-substituents, focused on phenyl-disubstituted quino-
lines (34–58). Whereas just one dimethoxy-substituted
compound with submicromolar efficacy against at least
one of the P. falciparum strains tested was identified
(22), three such dihalosubstituted compounds (42, 43,
and 45) were found, building upon reported results.6,7
Five such compounds were identified amongst the mono-
halo, monomethoxy-compounds prepared (50, 51, 55, 56,
and 58), structures for which there is no literature prece-
dence. In some halogen-containing compounds there
was little difference in activity between the chloro- and flu-
oro-analogs (comparison of 35 and 40, and 54 and 57), in
one case the fluoro-analog appears more active (38 and
44), whereas in other cases the chloro-analog appears
more active (36 and 42, 37 and 43, 39 and 45). The nine
submicromolar inhibitors included both 3- and 4-quino-
lines, in contrast to previous findings favoring 3-quino-
lines,4 and possessed either 2,5- or 3,4-disubstitution on
the 1-phenyl substituent. Preliminary pharmacophore
modeling using CATALYST software suggests that the
1-phenyl ring contributes as an aromatic hydrophobe
and one of the R1-substituents (either methoxy or halo)
as an aliphatic hydrophobe.
Compounds were prepared by the condensation of
substituted methyl ketones with substituted aldehydes
as shown in Scheme 1. The optimum conditions for this
condensation depended upon both the nature of the
aldehyde and the substituents on the ketone. For all
the naphthaldehyde-derived compounds, standard con-
ditions of sodium hydroxide in aqueous ethanol (Meth-
od A)9 gave good yields. For condensations involving
quinoline carboxaldehydes, barium hydroxide in metha-
nol (Method B) was used, since Method A typically
failed to provide the desired propenones. For a small
number of reactions involving dihalogenated methyl
ketones with quinoline carboxaldehydes, both Methods
A and B did not furnish the desired propenones. Usually
in these cases condensation could be mediated using
diethylamine in pyridine (Method C).10 In a few instanc-
es, mainly involving 2-quinolines, none of the methods
described provided the desired product. The structures
of the compounds prepared, and results from their bio-
logical testing, are shown in Table 1. All compounds
were assayed in vitro against P. falciparum, and selected
compounds in a P. berghei-infected mice model and for
predicted metabolic stability, using protocols that have
been reported previously.11–13
Most of the compounds were active in vitro against both
strains of P. falciparum tested (IC50 < 10 lM). One
determination of each IC50 was made; in other studies,
when multiple determinations were made standard devi-
ations were generally small, such that a fourfold differ-
ence in activity is likely to be significant. Efficacies in
the chloroquine-, quinine-, and pyrimethamine-resistant
W2 strain and in the mefloquine-resistant D6 strain were
similar, a preliminary indication that the compounds are
devoid of significant cross-resistance (resistance indices
ꢀ1). Both the 3-aryl and the 1-phenyl substituents
(R1- and R2-, respectively) impact activity, as reported
O
O
1
H
a
R2
+
R1
R1
O
R2
3
2
Scheme 1. Reagents and conditions: (a) excess 50% aq NaOH, EtOH
(Method A) or 1.2 equiv Ba(OH)2ꢁ8H2O, MeOH (Method B) or excess
Et2NH, pyridine (Method C).
Of these nine most active compounds, five were selected
for in vivo efficacy and in vitro metabolic stability
Gutteridge, Clare E.
