to introduce phenyl telluride (and other phenyl chalcogenides)
into molecules via IMCRs. Secondly, the synthetically more
accessible acid, aldehyde, and amine building blocks remain
variable and can be used to contribute further functionalities
to the multicomponent reaction product.
compounds 10 and 11 being the most active. These findings
are in excellent agreement with previous studies emphasizing
the activity of tellurium agents in this assay. In order to rule
out any major antioxidant activity (which may be counter-
productive), the thiobarbituric acid (TBA) assay has also been
performed (see ESIw for details). This assay measures the
ability of compounds to sequester oxygen-based radicals.
Most compounds were not particularly active in this assay.
Only the tellurium compounds 10 and 11 showed a significant
activity, probably due to the higher reducing power of tellurides
compared to selenides. Interestingly, HRMS also gave signals
for telluroxides, but not selenoxides, pointing towards a ready
oxidation of the tellurium compounds. Most compounds
complied with Lipinski’s rules of five regarding hydrogen
donors and acceptors, and were only slightly over a molecular
mass of 500 (9 being an exception). Importantly, partition
coefficients logPOW were in the range of 1.40 to 4.23, i.e. above
À0.4 and well below 5 (see ESIw).
This allows us to generate selenium and tellurium containing
molecules, which combine two, three or even four redox sites
in one molecule at will. A selection of products is shown in
Fig. 2, with Table 1 providing an overview of building blocks
used, reaction conditions, and yields obtained (details of the
synthesis and analytical data are provided in the ESIw). One
may note the mild reaction conditions employed (water as
solvent, room temperature) and the good yields obtained
(e.g. up to 96% for tellurium compound 10).11 Both aspects
of the IMCR compare highly favorably to conventional
methods employed in selenium and tellurium synthesis, such
as the ones used previously to synthesize quinone–chalcogen
redox agents.1 The synthetic approach described here together
with the availability of a large and diverse arsenal of redox-
active acids, aldehydes, isonitriles and amines—many of which
are suitable building blocks—enables the design and synthesis
of an unprecedented range of highly functionalized redox
agents. This may ultimately also allow QSAR-relationships,
for instance for cancer-type selective targeting.
Based on the rather promising estimates obtained in vitro,
some of the compounds were studied in cell culture. Here, we
briefly present and discuss the results for compounds 5, 6 and
8 obtained in three rather distinct assays which in combination
provide some information regarding cytotoxicity and selectivity:
firstly, an IC50 determination in cell culture employing cancer-
like, permanently growing L-929 murine fibroblast cells and
A-431 human epidermal tumor cells—both of which are
commonly used to screen for cytotoxicity against mammalian
cells. Secondly, a toxicity screen in cultured PC-3 human
prostate cancer cells. Thirdly, a single- and five-dose screen
in 58 tumor cell lines clustered in cells representing leukemia,
non-small cell lung cancer, colon cancer, cancer of the central
nervous system, melanoma, ovarian cancer, renal cancer,
prostate cancer and breast cancer, performed independently
by the National Cancer Institute (NCI) of the National
Institute of Health (NIH) in Bethesda, MD (US), to estimate
selectivity and to identify possible cancer targets. The results
obtained in L-929 cells are presented as example in Fig. 3; the
other cell culture results are provided in the ESIw.
The comparably straightforward synthesis of such multi-
functional agents does not, of course, address the question if
such rather large and complex agents are at all useful in cancer
therapy. We have therefore performed the thiophenol assay,
which measures catalytic activity of compounds in the pre-
sence of thiols and H2O2. This assay has been used as a
predictor of activity in cell culture.1 The results shown in
Table 1 confirm that all compounds enhance the oxidation of
thiols in the presence of H2O2. Several compounds are even
considerably more active than the benchmark compound
ebselen (1.5-fold increase vs. DMSO), with tellurium
Interestingly, compounds 6 and 8 showed promising results
in the L-929, A-431 and PC-3 assays as well as in the 58 cell
line screen whereas 5 was less active. In the L-929 cell assay,
8 killed these cells rather efficiently with an IC50 value of
0.7 mM (Fig. 3). A comparable IC50 value was obtained for the
Passerini diquinone 6 (1.1 mM), while 5 was considerably less
active (see Fig. 3). 8 was also rather active in the A-431 cell
line, with an IC50 of 5 mM. 5 and 6 were less active in this cell
line (IC50 of 135 mM and 55 mM, respectively). In PC-3 cells,
concentrations of 10 nM of 5, 6 and 8 reduced cell survival to
approximately 80%, while a concentration of 10 mM was quite
toxic, e.g. reducing survival to 10% for compound 8 (see ESIw).
These numbers are promising but do not indicate if this is
based on a general cytotoxicity, or if these compounds possess
selectivity against certain types of cancer cells which may be
useful for future therapy. These questions were addressed by
the 58 cell line screen. Compounds 6 and 8 were most active in
the 10 mM one-dose screen and were selected for five-dose
testing (for results see ESIw). In contrast, quinone-sulfide 5
was less active. LC50 values for 6 and 8 in the 58 cell lines were
generally around 1–10 mM, which is in good agreement with
Fig. 2 Multifunctional redox agents synthesized employing the
Passerini and Ugi reactions. Experimental details are provided in the
text, in Table 1 and as part of the ESIw.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 4702–4704 | 4703
Shabaan, Saad
