Abstract
A semiempirical molecular-orbital method (CAChe 4.9, PM5) was applied to delineate the relationship between the cytotoxicity (evaluated by 50% cytotoxic concentration, CC50) of twelve dihydroimidazole derivatives, their molecular weight and the eleven chemical parameters (descriptors) determined by CONFLEX/PM5 method. There was no correlation between the CC50 value and heat of formation, dipole moment, EHOMO, ELUMO, absolute hardness (η), absolute electron negativity (χ), reaction index (ω), length of substituted group in all cell lines tested. However, there was good correlation between the CC50 and the log P, or molecular size in all cell lines used. Surface and volume calculated by this method are useful in evaluating the cytotoxicity of dihydroimidazole derivatives.
We have previously reported the cytotoxic activity (1), and the quantitative structure-activity relationship (QSAR) between the cytotoxicity and twelve chemical parameters (descriptors) (2) of 4-trifluoromethylimidazole derivatives. Imidazole derivatives are well known to inhibit drug-metabolizing enzymes via direct binding to heme-Fe, and thus enhance the pharmacological actions of the concomitantly administered drugs (3). They are also known as inhibitors of p38 mitogen-activated protein kinase, anti-inflammatory agents, angiotensin II receptor antagonist, fungicides and herbicides (4). Some imidazole derivatives have been reported to induce apoptosis in Ehrlich ascites tumor cells, via a mechanism involving the activation of a pro-apoptotic protein BAK and a caspase-activated DNase (5). The biological activities of these types of compounds are generally affected by the substituted groups of the backbone structure. As an extension of our search for tumor-selective compounds that show higher cytotoxicity against human oral squamous cell carcinoma (OSCC) cell lines, we performed similar QSAR analysis of newly synthesized imidazole derivatives. For this purpose, we first determined their 50% cytotoxic concentration (CC50) against human promyelocytic leukemia (HL-60) and human OSCC cell lines (Ca9-22, NA, HSC-2, HSC-3, HSC-4), and evaluated the correlation between their molecular weight and eleven chemical descriptors determined by CONFLEX/PM5 method.
Materials and Methods
Materials. The following chemicals and reagents were obtained from the indicated companies: Dulbecco's modified Eagle's medium (DMEM) (Gibco BRL, Grand Island, NY, USA); fetal bovine serum (FBS) (JRH Bioscience, Lenexa, KS, USA); 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma Chem. Co., St. Louis, MO, USA). All dihydroimidazoles were provided by Dr. Kawase, Matsuyama University.
Assay for cytotoxicity. Human promyelocytic leukemic cell line (HL-60) (supplied by Professor Nakaya, Showa University) and human oral squamous cell carcinoma cell lines (Ca9-22, NA, HSC-2, HSC-3, HSC-4) (supplied by Professor Nagumo, Showa University) were cultured in RPMI-1640 or DMEM supplemented with 10% heat-inactivated FBS under a humidified 5% CO2 atmosphere, respectively. These cells were incubated for 48 hours with different concentrations of each compound, and the viable cell number was determined by cell counting after trypan blue exclusion test (for HL-60 cells) or MTT method (for other cell lines) (1). The 50% cytotoxicity (CC50) against these cell lines was determined from the dose-response curve.
Calculation. The most stable configuration of twelve dihydroimidazoles was calculated by CONFLEX 5 (Conflex Co. Ltd., Tokyo, Japan) (Figure 1). The optimization of the structure was achieved using a semiempirical molecular-orbital method (PM5), using a CAChe Worksystem 4.9 (MOPAC, PM5, non-COSMO, COSMO) (Fujitsu Co. Ltd., Tokyo) to first determine the heat of formation (COSMO, non-COSMO; kcal/mole) (Figure 2). From the data of heat of formation, the following descriptors were delineated: dipole moment (D), hydrophobicity (log P), highest occupied molecular orbital energy (EHOMO; eV), lowest unoccupied molecular orbital energy (ELUMO; eV), absolute hardness [η=(ELUMO—EHOMO)/2; eV)], absolute electron negativity [χ=—(ELUMO + EHOMO)/2; eV] and reactivity index (ω=χ2/2η; eV). Other descriptors such as surface area (Å2), volume (Å3), maximum length of substituted group at R1, R2 or R3 (Å) and of the molecule (Å), and the non-descriptor (molecular weight) were also calculated. The QSAR was performed with each descriptor (determined from molecular structure) and CC50 value (plotted as logarithmic scale), using a CAChe Worksystem 4.9 project reader.
The structure of dihydroimidazole derivatives.
Results
Table I shows the CC50 value (experimental values) of twelve dihydroimidazole derivatives against HL-60, Ca9-22, NA, HSC-2, HSC-3 and HSC-4 cells, and the chemical descriptors (calculated values) such as heat of formation, dipole moment, log P, EHOMO, ELUMO, η, χ, surface area, volume, maximum length of substituted group at R1, R2 or R3 and the maximum length of the molecule, and molecular weight. In contrast to 4-trifluoromethylimidazoles (1), the dihydroimidazole compounds used here showed much weaker cytotoxicity. We, however, dared to perform QSAR with the intention of designing more active compounds.
From these data, QSAR between CC50 and each descriptor was delineated (Figure 3). The correlation coefficients for the regression between CC50 and chemical descriptors (calculated in COSMO) are shown in Table II. There was no significant correlation between the CC50 and heat of formation, dipole moment, EHOMO, ELUMO, η, χ, ω or the maximum length of substituted group at R1, R2 or R3 (Table II). On the other hand, there was good correlation between the CC50 and molecular size as represented by the surface area, the volume and the maximum length of the molecule (Table II). There was some correlation between the CC50 and hydrophobicity. There was some correlation between the CC50 and molecular weight only in HL-60 cells.
Discussion
We have previously reported the lack of correlation between the CC50 and heat of formation, dipole moment, EHOMO and η of 4-trifluoromethylimidazole (2), consistent with QSAR of dihydroimidazoles used in the present study.
We found that the cytotoxicity of dihydroimidazoles highly correlated with the molecular size. We have not investigated the correlation of these factors with 4-trifluoromethylimidazoles, and therefore we performed this in the present report. As expected, we found similar good correlation between the cytotoxicity and surface area and volume of 4-trifluoromethylimidazoles in HSC-3 cells (Figure 3 f, g).
In conclusion, the present QSAR analysis of newly synthesized dihydroimidazoles demonstrates the lack of correlation between the CC50 value and heat of formation, dipole moment, EHOMO, ELUMO, η, χ, ω, length of substituted group in all cell lines tested. However, there was good correlation between the CC50 and the log P, and molecular size in all cell lines used. Surface and volume calculated by this method are useful in evaluating the cytotoxicity of dihydroimidazole derivatives.
The most stable conformation of dihydroimidazole derivatives used.
Correlation between CC50 value (log scale) and each descriptor of dihydroimidazoles against HL-60 cells (a-e), and of 4-trifluoromethylimidazoles against HSC-3 cells (f, g). The investigated descriptors are (a) hydrophobicity (log P); (b, f) surface area; (c, g) volume; (d) maximum length; (e) MW.
CC50 and chemical descriptors for dihydroimidazole derivatives.
Correlation coefficients between CC50 and each chemical descriptor in four different cell lines.
- Received July 6, 2009.
- Revision received October 29, 2009.
- Accepted November 5, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.