H. Zhang, et al.
Bioorganic&MedicinalChemistryLetters30(2020)127217
Fig. 1. Structures of fexinidazole (1) and previously reported covalent inhibitor of rhodesain (2).5
the antitrypanosomal activity of the analogues and they are generally
less selective (T. brucei/Hep G2) than 2 (Table 1). Also, compounds 3
compounds with propyl and butyl side chains, 5 and 6, have compar-
able inhibitory activity against rhodesain as 2 (Table 1).
for the treatment of visceral leishmaniasis by DNDi.9 The susceptibility
of pathogenic protozoans such as T. brucei, T. cruzi and Leishmania spp.
to nitroaromatics is linked to selective bioactivation by type I nitror-
eductase.10–13 It is worth pointing out, however, that despite the clin-
ical use of nifurtimox, benznidazole and fexinidazole in the treatment of
Chagas disease and HAT, concerns about the inherent mutagenic po-
tential of nitro groups remain valid and require careful evaluation.14
Also, development of resistance to nifurtimox and benznidazole by
T. cruzi, and the discovery of cross-resistance to fexinidazole by ni-
furtimox-resistant T. brucei has led to suggestions that new anti-
trypanosomal nitroaromatics should only be used as one of the active
ingredients in combination therapies.8,15–19 Like combination thera-
pies, dual-acting agents also present unique opportunities and chal-
lenges.20,21 The potential dual-acting (inhibition of trypanosomal ca-
thepsin L and trypanocidal action of the nitroaromatic moiety) nature
of compounds 16 and 17 provide a unique opportunity to deliver two
warheads in the same molecule. Dual-acting nitroaromatic compounds
can potentially suppress or delay development of drug resistance to
nitroaromatic-induced trypanocidal activity. It is important to point out
that dual-acting compounds can potentially have a higher degree of off-
target toxicity. The possibility that compounds 16 and 17 could act as
dual-acting agents needs validation. Also, the compounds require
multiparameter optimization of selectivity and pharmacokinetic prop-
erties. Boechat and co-workers have already shown that mutagenicity
of the nitro group on the imidazole ring.22 Therefore, the potential
mutagenicity of nitroaromatic-based dual-acting compounds can con-
ceivably be investigated and eliminated or minimized through sys-
tematic structure-activity relationship studies.
The results from rhodesain inhibition studies are consistent with
previous work on analogues of K777, as inhibitors of cruzain and rho-
desain, by Renslo and co-workers.6 Renslo and co-workers previously
reported that replacing a phenethyl group with a butyl group have very
marginal impact on both enzyme inhibitory and antiparasitic activities.
In this work, however, when the compounds’ activities on trypano-
somes and rhodesain are taken into consideration, the phenethyl side
chain is the most selective (T. brucei/Hep G2) motif out of the six motifs
investigated. Therefore, the phenethyl group was retained in our sub-
sequent SAR studies.
The quinoline ring was then replaced with quinoxaline (8), as well
as partially saturated quinoline derivatives such as methylated dihy-
droquinoline (9), methylated tetrahydroquinoline (10) and tetra-
hydroquinoline (11) rings. The compound bearing the quinoxaline
motif, 8, had similar inhibitory effects on T. brucei and rhodesain as 2
but it is less selective when compared to Hep G2 cells. However,
compounds bearing the partially saturated and methylated dihy-
droquinoline and tetrahydroquinoline motifs (9 and 10) are more se-
lective and have marginally higher activities against T. brucei.
Paradoxically, compounds 9 and 10 have much weaker/slower in-
hibitory activity towards rhodesain (Table 1). Perhaps, the relative
flexibility and bulkiness of the methylated and partially saturated rings
in dihydroquinoline (9) and tetrahydroquinoline (10) provides selec-
tive binding to its target in T. brucei. In addition, it is possible that the
partially saturated quinoline motifs in 9 and 10 are intrinsically less
toxic than the bicyclic aromatic quinoline ring, as previously suggested
In order to understand the key structural features important for the
inhibition of rhodesain by compounds that showed significant inhibi-
tion of the protease like 11 and 17 and those with much weaker in-
hibition like 9 and 10, covalent docking was used to predict the top
binding poses of each compound. As shown in Figs. 2 and 3 below, the
site and P1ʹ site, respectively. This is consistent with the orientations of
by Brinen and co-workers.23 The dihydroquinoline moiety in 9 and the
tetrahydroquinoline moiety in 10 and 11 were predicted to bridge the
P2 and P3 sites.
To further explore this series of vinyl sulfone-based compounds, the
quinoline ring in 2 was replaced with monocyclic heterocycles (12–17).
As shown in Table 1, the pyridine-based analog, 12, was inactive
against T. brucei, while the 1,4-dimethyl-1H-imidazole-based analog,
13, had much weaker activity against T. brucei. The 1,3-thiazolidine-
based compound, 14, had similar potency against trypanosomes as 2,
but compounds 12–14 are generally less selective. Nevertheless, we
found out that nitroaromatic-based compounds 16 and 17 are sig-
nificantly more potent against T. brucei than 2, and they both have an
adequate degree of selectivity. It is well-established that protozoan
parasites like trypanosomes and Leishmania are susceptible to ni-
troaromatics. In fact, the front-line therapies to treat HAT includes ni-
troaromatics such as nifurtimox, used in combination with eflornithine,
and the recently approved fexinidazole (1).8 Benznidazole, a ni-
troimidazole, is the front-line drug against Chagas disease. Another
nitroimidazole derivative, DNDI-0690, is under clinical development
The nitrofuran ring in 17 is predicted to occupy the P2 site and it is
solvent exposed. When compared with 9 and 10, compound 11 is a
better inhibitor of rhodesain. It is likely that the three methyl sub-
stituents on the partially saturated quinoline ring precludes efficient
binding of 9 and 10 to rhodesain when compared to 11. Also, the hy-
drophobicity of solvent-exposed methyl groups in 9 and 10 possibly
negates the formation of solvent-stabilized rhodesain-inhibitor com-
plex.
2
Ajayi, Oluwatomi
