D. Sato et al. / International Journal of Antimicrobial Agents 35 (2010) 56–61
57
shown to be highly toxic to bacteria, Porphyromonas gingivalis, T.
vaginalis and E. histolytica in vitro [6–9]. It was also reported to
treat infections by P. gingivalis and T. vaginalis in a rodent model
[7,8]. In the present study, the in vitro and in vivo efficacy of
TFM and its derivatives were examined to gain insight into the
structure–activity relationship and to determine the mechanism
underlying their mode of action.
0.2 mol/animal) or orally through a stomach catheter (0.5 mL,
0.2–1.0 mol/animal) at 24 h after infection. Animals were sac-
rificed 6 days post infection and the liver and abscesses were
dissected and weighed separately.
3. Results and discussion
The in vitro cytotoxicities of TFM, 15 TFM derivatives and
4 control compounds lacking fluorine were evaluated by mea-
suring their 50% inhibitory concentration (IC50) values against
E. histolytica trophozoites (Table 1). Whilst TFM showed an
IC50 of 7.34 M, the derivatives TFM-01 and TFM-02 had
IC50 values of 2.22–2.28 M. Three difluoroanilide compounds
reduction in their IC50 values compared with TFM. Three deriva-
tives with 2,5- and 3,4-dimethoxyanilide (TFM-07 and TFM-08)
or 3,4,5-trimethoxyanilide (TFM-09) modification had a further 2-
fold improvement in their IC50 values. These values are 1.6–4.3-fold
lower than metronidazole (mean standard deviation in triplicate,
4.76 M 0.22). In contrast, compounds containing additional
functional groups on the phenyl ring of TFM-01 [2-methyl-
4-chloroanilide (TFM-10), 2-methoxy-4-bromoanilide (TFM-11),
2-methoxyl-4-chloro-5-methylaniline (TFM-12)] or bulkier struc-
tures such as 8-aminoquinoline (TFM-13), 5,6,7,8,-tetrahydro-
1-naphthalenamine (TFM-14) and 4-bromo-naphthalenamide
(TFM-15) had higher IC50 values compared with TFM. The IC50 val-
ues of TFM-10 and TFM-15 were >80 M, indicating a loss of efficacy
in these derivatives. Furthermore, structurally similar derivatives,
in which the trifluoromethyl group was substituted by a methyl
group (MET-01 or MET-09), failed to produce cytotoxic activity
(IC50 >80 M). These results indicate that the trifluoromethyl group
at the C5-carbon of TFM and its derivatives is essential for their
cytotoxicity.
2.1. Chemical synthesis of trifluoromethionine and its derivatives
The synthetic scheme and structure of TFM and its analogues
are shown in Supplementary Fig. 1.
2.2. Parasites and cultivation
Trophozoites of E. histolytica HM-1:IMSS cl6 were cultured
axenically in BIS medium at 35.5 ◦C as described previously [10].
2.3. In vitro cytotoxicity assay of trifluoromethionine derivatives
against Entamoeba histolytica trophozoites and Chinese hamster
ovary (CHO) cells
Approximately 5 × 103 amoebae in 280 L of medium were
seeded into each well of a 96-well plate and incubated with various
concentrations of TFM derivatives for 48 h. Following incubation,
the amoebae were washed with 100 L of pre-warmed Opti-MEM®
I (Invitrogen, Carlsbad, CA) and viable cells were counted using
Cell Proliferation Reagent WST-1 (Roche Diagnostics, Mannheim,
Germany). Cytotoxicity against mammalian cells was examined as
described above except that 1 × 104 CHO cells were cultivated in
280 L of F-12 medium (Invitrogen) under 5% CO2 at 37 ◦C.
To determine whether ␣-keto acids generated by the release
of trifluoromethane thiol from the TFM derivatives is responsible
for their cytotoxic activity, OB-01 and OB-09, which were predicted
products of TFM-01 and TFM-09, respectively, were evaluated. Nei-
ther OB-01 nor OB-09 showed cytotoxic activity (IC50 >80 M).
These results confirmed that the trifluoromethyl group is respon-
sible for the cytotoxicity of TFM and its derivatives [7–9].
2.4. Assay for l-methionine ꢀ-lyase activity
Recombinant MGL1 (3 g) or MGL2 (1.2 g) (final concen-
trations 15 g/mL and 6 g/mL, respectively) [5] was incubated
200 L of 100 mM sodium phosphate buffer (pH 7.2) containing
20 M pyridoxal-5ꢀ-phosphate (PLP) and 2.5 mM 5,5ꢀ-dithiobis(2-
nitrobenzoic acid) (Sigma, St Louis, MO) at 37 ◦C. Released thiols
were measured as described previously [9].
To assess whether TFM derivatives were selective towards E.
histolytica, the efficacy of TFM, TFM-01 and TFM-02 against a rep-
resentative mammalian cell line was examined. The IC50 values
of TFM, TFM-01 and TFM-02 against CHO cells were more than
100-times higher than for E. histolytica (709 172, 982 174 and
dimethoxyanilide derivatives were higher or slightly lower (>1 mM
and 554 94 M, respectively). It was also previously reported that
a high concentration of TFM (100 g/mL, corresponding to 493 M)
inhibited the growth of mouse myeloma cells [8]. Taken together,
TFM and its derivatives have good selectivity toward E. histolytica.
To determine whether the observed increase in the cytotoxic
effects of the amidated TFM derivatives compared with TFM was
due to their higher degradation efficiencies by MGL, time kinetics of
the degradation of TFM, TFM-01 and TFM-02 by two recombinant
MGL proteins (MGL1 and MGL2) were investigated [5]. Trifluo-
romethane thiol generated from TFM by MGL1 or MGL2 increased
with incubation time (Fig. 1). Degradation of TFM by MGL2 was
ca. 12.5-fold faster compared with MGL1 at 10 min after incuba-
tion, which is also consistent with a previous study [5]. In contrast,
decomposition of TFM-01 and TFM-02 by MGL1 and of TFM-02 by
MGL2 was negligible, whilst that of TFM-01 by MGL2 was ca. 70%
slower compared with TFM.
2.5. Degradation of trifluoromethionine derivatives in amoebic
cell lysate determined by capillary electrophoresis with
electrospray ionisation time-of-flight mass spectrometry
(CE-TOFMS)
Approximately 3 × 106 trophozoites were lysed with 450 L
of 10 mM HEPES (pH 7.5) containing 1 M PLP and 0.05% Tween
TFM-01, OB-01 (see Supplementary Fig. 1) or dimethyl sulphox-
ide (DMSO) at 37 ◦C for 2 h. The reaction was terminated by adding
10 times volume of methanol. TFM and 2-oxobutyrate were quan-
tified by CE-TOFMS [11] in cationic and anionic mode, respectively,
under conditions described previously [11]. TFM and sodium 2-
oxobutyrate were used as standards.
2.6. Evaluation of the amoebicidal activity of trifluoromethionine
derivatives using a hamster liver abscess model
Approximately 1 × 106 trophozoites were injected into the left
lobe of the liver of 2–3-week-old Syrian hamsters (mean standard
error body weight, 39.9 0.56 g). TFM or its derivatives dis-
solved in DMSO were administered either subcutaneously (0.1 mL,
We further examined the degradation process of TFM-01 in
the amoeba by directly measuring the products formed when