Q. Yang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4602–4607
4605
Figure 3. Western blots for the analysis of caspase-3 activation/cleavage, PARP cleavage, and p53 expression in A2780 ovarian cancer cells after 24 h treatment with (R,R)-5a
at various concentrations (2.5, 5, and 10 M). b-Actin was used as a loading control. The term ‘c’ refers to untreated control cells. (A) Relative quantity of cleaved caspase-3
l
relative to b-actin. (B) Relative quantity of cleaved-PARP relative to b-actin. (C) Relative quantity of p53 relative to b-actin. The data represent the mean SD of three
independent experiments.
asymmetric reductive homocoupling of the corresponding (S)-2a
and (S)-2b mediated by SmI2 in THF at ꢀ78 °C in the presence of
HMPA; these products were produced at yields of 66% and 74%,
respectively. Finally, (1R,2R)-3a and (1R,2R)-3b were converted to
their corresponding enantiopure free diamines (R,R)-4a and (R,R)-
4b under acidic conditions at yields of 85% and 94%, respectively.
Using (R)-2-methyl-2-propanesulfinamide [(R)-1] as the starting
material, both (1S,2S)-4a and (1S,2S)-4b were synthesized in the
same manner as (R,R)-4. The enantiomeric purity of the two pairs
of diamines 4a and 4b were then ascertained by chiral HPLC anal-
ysis of the corresponding diacetate.32 The diacetates of diamines
(R,R)-4a, (S,S)-4a, (R,R)-4b, and (S,S)-4b each showed an extremely
high ee of >99%.
With the enantiopure diamines (R,R)-4 (a, b) and (S,S)-4 (a, b),
IrCl3ꢁ3H2O was used for the coordination to form the desired com-
plexes (Scheme 2). Due to the chemical inertness of IrCl3ꢁ3H2O, the
coordination reaction was conducted under rather harsh condi-
tions, i.e., high temperatures and prolonged reaction times, accord-
ing to the modified method described by Galsbol et al.33 These
reactions resulted in the formation of the final iridium(III) com-
plexes (R,R)-5 (a, b) and (S,S)-5 (a, b) at yields in the range of
28–39%.34 All of the synthesized iridium(III) complexes were char-
acterized by IR, 1H NMR, 13C NMR, and mass spectra (ESI and HRES-
I). The spectroscopic and analytical data of these iridium(III)
complexes were in complete agreement with their assigned struc-
tures. In the IR spectra, the bands of the N–H stretching vibrations
of the complexes 5 were redshifted relative to the single amino
group of the corresponding ligands due to the coordination of the
amino group with iridium. The positive ESI mass spectra of all of
the iridium(III) complexes (R,R)-5 (a, b) and (S,S)-5 (a, b) showed
the expected [M–Cl]+ peaks, each of which had six ion peaks due
to the presence of iridium and chlorine isotopes. Since the stable
five membered chelate ring at the iridium blocks rotation around
the C–N axis, both N-bound protons become diastereotopic owing
to the neighborhood of the asymmetric C-atoms. This led to the
appearance of separate signals for the axially and for the equatori-
ally orientated N–H atoms in the 1H NMR spectra of these irid-
ium(III) complexes. Furthermore, the structure of compound
(R,R)-5a was unambiguously confirmed by single crystal X-ray
structure determination (Fig. 1).35 The two chloride ligands of
(R,R)-5a were found to be situated oppositely, and the two
asymmetric C-atoms of the 1,2-di(4-methylphenyl)ethane-1,2-dia-
mine ligands possess the (R,R)-configuration.
To evaluate their anticancer activity, the human solid tumor cell
lines A2780 (human ovarian carcinoma), A549 (human non-small
cell lung cancer), KB (human oral epithelial carcinoma), and
MDA-MB-231 (human breast cancer) were used. The in vitro cyto-
toxicity was evaluated after the cells were exposed to the irid-
ium(III) complexes 5 for 72 h using the MTT assay.36 Oxaliplatin
was used as the positive control. The results, which are expressed
as IC50 values (drug concentration giving 50% survival),37 are
shown in Table 1.
For the p-methyl-substituted complexes 5a and the p-trifluoro-
methyl-substituted complexes 5b, a pronounced difference in the
activities of the (R,R)- and the (S,S)-isomers was observed on all
tested tumor cell lines. Both (R,R)-configured complexes exhibited
significantly higher activity than their corresponding (S,S)-enantio-
mers. The activities of (R,R)-5a and (R,R)-5b were higher than or at
least comparable to that of oxaliplatin, whereas (S,S)-5a and (S,S)-
5b were almost inactive. Complex (R,R)-5a had lower IC50 values
than oxaliplatin toward A2780, KB, and MDA-MB-231 cells, and
complex (R,R)-5b exhibited better activity than oxaliplatin against
A2780 and MDA-MB-231 cells. The antitumor activity of these
complexes toward A2780 and MDA-MB-231 cells decreased in
the following order: (R,R)-5a > (R,R)-5b > oxaliplatin. For the KB
cells, the order of cytotoxicity was (R,R)-5a > oxaliplatin > (R,R)-
5b. The analysis of the A549 cells revealed that neither (R,R)-5a
nor (R,R)-5b exhibited higher activity than oxaliplatin (ꢂtwo-fold
less potent). The above results indicate that the antitumor activity
of these iridium(III) complexes may depend, in addition to the crit-
ical configuration of the 1,2- diarylethane-1,2-diamine moiety, on
the type of tumor cells and the nature of the substituents on the
phenyl rings.
Due to their promising potent cytotoxicity, it was of interest to
further explore the molecular mechanism underlying the (R,R)-5-
induced cell death. To investigate whether the (R,R)-5-initiated cell
death is mediated by apoptosis, A2780 cells were treated with
(R,R)-5a or oxaliplatin for 72 h and then analyzed by flow cytome-
try through Annexin V/PI staining. The apoptotic rates, including
early and late apoptosis, for A2780 cells treated with 5, 10, and
20
lM (R,R)-5a were 31.8%, 52.1%, and 69.1%, respectively, whereas
the apoptosis rates for A2780 cells untreated or treated with 20
lM