X. Wang, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxxx
Table 2
Physicochemical and in vitro ADME properties of 2 and its most
potent analogs 5 and 19 are shown in Table 4. The gLogD7.4 values
ranging from 95 and 106 Å2, aq. solubilities ranging from 12.5 to >
100 µg/mL, and plasma protein binding < 90% suggest that these
compounds would be expected to have relatively good biopharmaceu-
tical properties. However, all of the compounds had relatively high
intrinsic clearance values in human and mouse liver microsomes in-
dicating that each would be expected to be rapidly metabolized. Me-
tabolic stabilities were higher in human vs. mouse liver microsomes.
Compound 19, the most lipophilic of the three, had the lowest meta-
bolic stability.
Cytotoxicity screen of 2, 5, and 19.
Compd
Cytotoxicity IC50 (µM)
HFF
U-2 OS
HEK 293 T
HC-04
2
> 50
2.0
0.15
0.16
28
0.36
0.95
42
> 50
> 50
> 50
5
19
> 50
Table 3
Activity of 2, 5, and 19 against a panel of protozoan parasites.
In this SAR scoping study, we started with 2, a known7 GLUT1 in-
hibitor. As we did not assess inhibition of glucose transport in T. cruzi17
by 2–19, we cannot confirm that the T. cruzi hexose transporter is a
potential cellular target for these compounds. In this respect, a recent
study12 claimed that nicotinamide phosphoribosyltransferase
(NAMPT), not GLUT-1, is the target of 1 and 2 in cancer cell lines. In
addition, posaconazole and a high proportion of hits identified in HTS
campaigns against T. cruzi appear to inhibit sterol 14α-demethylase (T.
cruzi CYP51), and most of these hits contain imidazole or pyridyl sub-
structures.1,18–21 Of the pyridyl-containing T. cruzi CYP51 inhibitors, a
sterically unhindered 3-pyridyl substructure was predominant. Thus,
we wondered whether T. cruzi CYP51 might be a potential target of this
compound series based on our observed SAR – particularly the dramatic
(three orders of magnitude) loss of potency for 14–16 when methyl
groups were introduced adjacent to the nitrogen atom in the 3-pyridyl
substructure of 2.
Compound
P. falciparum
T. b. rhodesiense
MHOM-ET
67/L82
NF54
STIB 900
RH-dTom
2
10
32
> 200
180
> 100
> 100
21
5
21
25
19
5.8
3.4
49
0.0042
0.0070
0.99
0.80
a
b
Mean from n ≥ 2, individual measurements differed by less than 50%.
The compounds were tested against L. donovani amastigotes in an axenic
assay.
c
RH (Type I) strain T. gondii with inserted fluorescent transgene dimerized
Tomato (dTom).
d
Chloroquine for P. falciparum; melarsoprol for T. b. rhodesiense; miltefosine
for L. donovani; pyrimethamine for T. gondii.
To test this hypothesis, 2 and 14 were analyzed as T. cruzi CYP51
heme binding ligands22 and for inhibition of enzyme activity in a re-
constituted T. cruzi CYP51 activity assay.23 While 2 induced a moderate
type II spectral response in the Soret band (Kd 1.32 µM, Fig. 2a), 14 did
not cause any changes (not shown). The IC50 values, determined in a
30 min reconstituted enzyme reaction, were 49 and 487 µM, respec-
tively (Fig. 2b). These data suggest that CYP51 could be a target for 2
factor of CYP51),22 although the inhibition at the enzyme level (IC50 of
49 µM) does not seem sufficient to fully account for the potency at the
cellular level (IC50 of 7 nM). For comparison, posoconazole similarly
inhibits the growth of T. cruzi (IC50 of 5 nM), but has a T. cruzi CYP51 Kd
of 18 nM.22,24 Thus, the mechanism of action of 2 probably includes
another target(s) which is also suggested by the dramatic loss of po-
tency against T. cruzi observed with small variations in the N-methyl-
piperazine substructure – e.g. piperazine 3 and morpholine 5. Finally,
the potent activity of 2 on T. cruzi and the weak activity against the
other protozoa suggest that the target may be highly T. cruzi specific.
Table 4
Physicochemical properties and in vitro metabolic stability of 2, 5, and 19.
h/m CLint
Compd gLogD7.4
2
2.8
2.6
4.1
95.8
> 100/25–50
45/112
85.2
5
100.6
105.7
50–100/12.5–25 69/200
76.2
19
> 100/25–50
535/677
not assessed
a
LogD values were estimated by correlation of their chromatographic re-
tention properties using a modified gradient HPLC method adapted from
b
c
Calculated using ChemAxon JChem for Excel.
Compounds in DMSO were spiked into either pH 6.5 phosphate buffer or
0.01 M HCl (approx. pH 2.0) and analyzed by nephelometry14 to determine a
concentration range.
d
In vitro intrinsic clearance measured in human and mouse liver micro-
somes.
e
Plasma protein binding was estimated using a gradient HPLC method15
where the chromatographic retention on a human albumin column was com-
pared against the properties of standard compounds with known binding va-
lues.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Leishmania donovani and Toxoplasma gondii. Compound 5 was 10-fold
less active against P. falciparum and T. b. rhodesiense than it was against
T. cruzi (Tables 1 and 3) and it had weak to no activity against the other
19 against T. cruzi (Tables 1 and 3), P. falciparum and T. b. rhodesiense,
protozoans.
Acknowledgment
We acknowledge the U.S. National Institutes of Health (AI116723-
01, GM103427, and GM067871) for financial support. We acknowledge
the contributions of the Centre for Drug Candidate Optimisation,
Monash University for conducting the ADME studies.
3