C. S. Reid et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6806–6810
6809
containing 5-imidazolyl and 2-pyrrolyl rings were at least an order
of magnitude less active than the corresponding 2-pyridyl com-
pound, while the 4-(2-methylthiazolyl) compound retained the
in vitro (but not in vivo) activity of the corresponding 2-pyridyl
congener.9 Regarding the linker, a bis-AIA with a phenyl linker
was inactive against intracellular Leishmania, but bis-AIAs bearing
4,40-biphenyl and 4,40-diphenyl ether linkers were as potent as the
corresponding bis-AIAs containing a 4,40-diphenylfuran linker.10
Concerning substitution on the diphenylfuran linker, inclusion of
halogen, alkyl or alkoxy groups on the phenyl rings improves po-
tency compared to the diphenylfuran derivative lacking such sub-
stitutions.8,9 The alkoxy groups on the phenyl rings can be quite
large, as the cyclopentyloxy-containing compound DB1852 retains
mid-nanomolar in vitro potency against intracellular Leish-
mania,9,12 but the benzyloxy derivative was 3-fold and sixfold less
potent than DB1852 and DB766, respectively.9 Our present study
indicates that bis-AIAs bearing unsymmetrical substitutions on
the diphenylfuran linker retain the nanomolar antileishmanial po-
tency of their symmetrical bis-AIA counterparts (Table 1).
Of the considerable number of bis-AIAs possessing potent
in vitro activity against intracellular Leishmania, only derivatives
bearing relatively large alkoxy substituents on a diphenylfuran lin-
ker such as DB766, DB1852, and compound 10 are relatively well
tolerated in mice. In a previous report, we examined the in vivo
properties of both DB766 and DB745, a compound identical in
structure to DB766 except that the isopropoxy groups were re-
placed by smaller ethoxy groups.11 DB745 displayed in vitro and
in vivo antileishmanial efficacy similar to DB766 but was signifi-
cantly more toxic to mice, despite the fact that comparable doses
of these compounds resulted in significantly higher exposure for
DB766. Considering that the in vivo toxicity data for compounds
6a–6d, like DB745, fit our structure–toxicity relationship outlined
above, we do not believe that it is justified to sacrifice more ani-
mals to measure the pharmacokinetics of these toxic compounds.
Like the lead compound DB766, its regioisomer 10 displayed no
overt toxicity to mice when given ip at 30 mg/kg/day ꢁ 5, but
was less effective than DB766 at reducing the liver parasite burden
in L. donovani-infected mice (Fig. 3). Previous work showed that
micromolar levels of DB766 and related AIAs accumulate in the
livers of mice after a single dose.11,12 If this is also the case for
10, sufficient compound should be present in the liver to eradicate
the parasites (Fig. 2B). Within hepatic tissue, L. donovani reside
within Kupffer cells, the macrophages of the liver. Since Kupffer
cells comprise only 2% of the liver cell population23 and AIA levels
have not been measured in this cell type, it is possible that 10 and
other AIAs do not reach sufficient concentrations within the para-
sites to eliminate infection. Testing this hypothesis will require the
isolation of Kupffer cells from animals dosed with an AIA and the
determination of compound levels within these cells.
Figure 3. Efficacy of 10 in a murine model of visceral leishmaniasis. Compound 10
and the antileishmanial standard drug miltefosine were administered in five daily
doses to L. donovani-infected BALB/c mice as indicated in the text, then animals
were euthanized and liver parasitemia was determined microscopically. Data are
presented as the percentage reduction of liver parasitemia compared to infected,
untreated control animals. Bars and error bars indicate the means and standard
deviations, respectively, of groups of four animals. ⁄P <0.05 compared to untreated
control.
day for five days. This dosing regimen was thus selected for the
evaluation of 10 in a murine model of visceral leishmaniasis
(Fig. 3). Antileishmanial efficacy was assessed by established meth-
ods in which L. donovani-infected BALB/c mice were treated with
test compound or vehicle daily for five days starting one week post
infection and liver parasitemia was assessed microscopically two
weeks post infection.19 Compound 10 reduced the liver parasite
burden by 32%, whereas the standard antileishmanial drug miltefo-
sine decreased liver parasitemia by 96% when administered orally
at a dose of 10 mg/kg/day ꢁ 5 (Fig. 3). A previous study showed that
the lead AIA DB766 reduced liver parasitemia by 63% when given ip
at a dose of 30 mg/kg/day ꢁ 5.11
AIAs display outstanding in vitro activity against intracellular
Leishmania as demonstrated in previous work8–11 and as shown
in Table 1. We are unaware of another structural class of molecules
that contains as many compounds possessing nanomolar potency
against intracellular Leishmania, although phospholipid analogs,20
arylisoquinolinium salts,21 and cyanines18,22 also exhibit mid-
nanomolar in vitro potency against Leishmania within host macro-
phages. Given the range of bis-AIAs that have been prepared and
tested for activity against intracellular Leishmania in vitro, a de-
tailed structure–activity relationship can now be outlined for these
compounds (Fig. 4). Regarding terminal groups containing a six
membered aromatic ring, 2-pyridyl groups are preferred over phe-
nyl,8 2-pyrimidinyl, or 2-pyrazinyl groups,9 and substitution of the
2-pyridyl terminal group at the para position with a halogen (but
not a methoxy) group reduces potency.9 For AIAs with a 5-mem-
bered heteroaromatic ring as the terminal group, compounds
Despite the potency of the AIAs against intracellular Leishmania
and the favorable distribution of members of this class of com-
pounds, we have thus far been unable to identify an AIA with the
combination of safety and efficacy required to progress to preclinical
Unsymmetrical bis-AIA (R2
When R2 = R3 = o-i-Pr and
either ortho or meta to the
furan, compounds are
active and well tolerated
Compared to compounds with
= o-i-Pr, R3 = H, F, Me,
unsubstituted (R2
= = H)
R3
OMe) are active but toxic
linkers, potency is improved when
R3
R
2 = R3 = halogen, alkyl or alkoxy
R2
NH
NH
R1
O
Activity is optimal
R1
N
N
H
when R1 = 2-pyridyl or
H
4-(2-methylthiazole)
A phenyl linker abolishes activity; compounds
with unsubstituted diphenylfuran, biphenyl, or
diphenyl ether linkers are active but toxic
Figure 4. Structure–activity map for AIAs against intracellular Leishmania.