1 Introduction
Phenolic compounds and their derivatives are constituents of numerous
synthetic and natural chemicals such as drugs, cosmetics, dyes,
agricultural products, pesticides, food additives etc. Their properties
have been frequently studied at the molecular and cellular level.
Phenolic-based compounds provide an excellent opportunity to examine
their biological activities and to analyze corresponding mechanistic
details in a comparative fashion using the quantitative structure
activity relationship (QSAR) paradigm. It has been useful in elucidating
the mechanisms of chemical–biological interactions in various
biomolecules, particularly enzymes, membranes, organelles and
cells.[1]
Selassie et al .[2] studied the cytotoxicity
of a series of 69 simple phenols in a fast growing murine leukemia L1210
cell line. They developed a Quantitative Structure–Activity
Relationship (QSAR) model where the cytotoxicity of the substituted
phenol was evaluated as log(1/IC 50), withIC 50 being its concentration that inhibited cell
growth by 50%. These values were correlated with Brown-Hammett
parameters (σ +), the HOMO–LUMO energy
difference (L–H gap), the logarithmic n-octanol - water
partition coefficients (log P ) and homolytic bond dissociation
energies (BDE ). Phenols with electron withdrawing substituents
were excluded from the analyses. BDE values were calculated at
B3LYP/6-31G**//AM1 level of theory with effective core potentials used
for bromine- and iodine-substituted molecules. L-H gap values were
obtained by the semiempirical AM1 method only. Despite these
restrictions, the developed QSAR models granted surprisingly good
results with 17 – 27 outliers depending on the descriptors used. It
seems that the above mentioned cytotoxicity descriptors cannot produce
better results.
Alagona and Ghio[3] evaluated the antioxidant
activity of various sites of prenylated pterocarpans according to their
Cu2+ coordination ability using the metal ion affinity
(MIA ) values determined by B3LYP calculations. Spin density of
the cation upon ligand coordination became negligibly low, whereas the
ligand spin density approached 1. Thus, the ligand was oxidized to a
radical cation (Ligand•+), while Cu(II) was reduced to
Cu(I). In agreement with experimental investigations, the higher
antioxidant activity of individual compounds and their reaction sites
can be assigned to higher MIA values and higher reducing
character towards Cu(II). The Cu2+ ion serves as a
probe to estimate the extent of the electron density transfer only and
the optimized structures need not correspond to any real complex.
Mammino[4] analogously tested antioxidant ability
of various sites of hyperjovinol A. Another modification of the
above-mentioned method by including Laplacians at Cu-N bond critical
points (BCP) as the measure of the corresponding electron density
transfer has been used for both N centres of a series ofpara -phenylene diamine antioxidants.[5]
Other 1st row transition metal ions were also used in
the role of similar probes for the compounds with oxygen active sites.
B3LYP, M06-2X, BHandHLYP, MPWLYP1M and G96LYP methods were used to study
antioxidant properties of kanakugiol through its Cu(II) and Co(II)
coordination ability[6] as well as lucidone,
linderone and methyllinderone through their Fe(II) coordination
ability.[7] Subsequently the authors continued in
studying antioxidant properties of butein and
homobutein[8], chalcone derivatives kanakugiol and
pedicellin[9] by considering their Fe(II) and
Fe(III) coordination ability using the B3LYP method. A systematic B3LYP
study of a series of para-phenylene diamine antioxidants with the late
1st row transition metals in various spin
states[10] has shown that using these M(II) probes
yield very similar results of antioxidant effectiveness.
This method was used for the relative cytotoxicity estimation of various
sites of model rutile nanoparticles as well.[11]It was shown that the experimentally observed higher cytotoxicity of
rod-like nanoparticles in comparison with the spherical ones might be
explained by the higher electron density transfer to the interacting
cells (as indicated by the extent of the electron density transfer to a
Cu2+ probe), whereas MIA reflects the
reactivity of individual active sites. The aim of our recent study is
the application of this method to predict the cytotoxicity ofp -substituted phenols. Our results will be compared with standard
QSAR correlations used by Selassie et
al .[2]