Search

In this study, the cytotoxicities of a series of X-thiophenols vs rapidly growing mouse leukemia cells in vitro are determined. The resulting ID50 values are then used to formulate a quantitative structure−activity relationship, which is well-correlated by the Brown variation of the Hammett electronic parameter, σ-plus (σ⁺ ), such that Log 1/ID50 = −0.93 (±0.18) σ⁺ + 0.86 (±0.24) IH + 3.99 (±0.13). IH represents an indicator variable that calls attention to the unusual activity of halogens and pseudohalogens. In lieu of σ⁺, homolytic bond dissociation energies (BDE) are also used successfully to correlate the cellular cytotoxicities of thiophenols. The nature of substituent effects on cellular toxicity is examined, and they reveal that electron-releasing substituted thiophenols such as 4-amino thiophenol and the 4-alkoxy thiophenols are highly cytotoxic and effective at inhibiting cellular proliferation at physiological pH. On the other hand, electron-attracting substituted thiophenols such as the 4-cyano and 4-halogen analogues show a reduced ability to inhibit the cell growth of this cell line. Thus, there is a clear parallel between enhanced biological activity and electron releasing ability as measured by σ⁺ constants or BDE values. The susceptibility of the cellular interaction to electronic effects as delineated by the coefficient with the σ⁺ term (also called the Hammett ρ value) is high (−0.96), suggesting that substantial energetic assistance is provided by the substituents and that a weak initiating radical reactant such as superoxide radical may be involved. Previous cytotoxicity studies of a large diverse data set of X-phenols in this cell line and embryo cells have also revealed a more pronounced dependence on σ⁺ and ΔBDE. A comparison of reaction constants obtained from thiophenoxy radical formation reactions and phenoxy radical formation reactions in organic media suggests radical-mediated involvement in cell cytotoxicity. Such cells could be more vulnerable to the effects of reactive thiyl species on their metabolism and subsequent proliferation.

This report considers the importance of electronic effects in their role in the QSAR of chemical−biological interactions. The problem of accounting for polarizability effects in ligand−substrate interactions is discussed in terms of molecular polarizability (MR) and NVE (number of valence electrons) using additive values for valence electrons. The two approaches give essentially the same result in examples of frog nerve toxicity and examples of nerve toxicity with rabbits and cockroaches. The point is made that no matter how one approaches QSAR, electronic interactions must be considered if we are to begin to develop a science of chemical−biological interactions.

In this report we illustrate the importance of making an effort to ensure that all of the dependent variables in a QSAR are associated with one mechanism of action. If this cannot be established, it will not be possible to compare the QSAR with others acting on different biological systems. That is, such information cannot be used to develop a science of chemical−biological interactions. A study of the action of a large set of phenols acting on Tetrahymena pyriformis made by Cronin and Schultz is analyzed to illustrate the problem.

Not Available

Thiopurine methyltransferase (TPMT) catalyzes the S-methylation of thiopurine and thiopyrimidine drugs. If potent TPMT inhibitors were available, studies of the regulation and properties of this drug-metabolizing enzyme would be facilitated. Each of a series of benzoic acid derivatives tested was found to inhibit purified human kidney TPMT. Concentrations required to inhibit TPMT by 50% ranged from 20 microM for 3,4-dimethoxy-5-hydroxybenzoic acid to 2.1 mM for acetylsalicylic acid. Inhibition was noncompetitive or mixed with respect to both S-adenosyl-L-methionine, the methyl donor for the enzyme, and 6-mercaptopurine, the methyl acceptor substrate. Preliminary structure-activity relationship analysis demonstrated that the benzoic acid structure was important for inhibitory activity, and that inhibition was enhanced by the addition of methoxy and/or phenolic hydroxyl groups to the ring. Quantitative structure-activity relationship analysis performed with additional benzoic acid derivatives showed that inhibitory activity could be modeled well by an equation that included the normal Hammett constant and a parameter, pi', related to lipophilicity. Several nonheterocyclic aromatic thiol compounds, including thiophenol and thiosalicylic acid, were discovered to be substrates for TPMT. Apparent Km constants for some of these aromatic thiol compounds were in the nanomolar range, several orders of magnitude lower than those of the thiopurines and thiopyrimidines previously thought to be the only substrates for TPMT. These observations suggested that "aryl thiol methyltransferase" might be a better name than "thiopurine methyltransferase" for this enzyme. Discovery of new classes of inhibitors and substrates for this important drug-metabolizing enzyme has implications for drug metabolism research and for clinical medicine.

The mutagenicities of five heterocyclic 3,3-dimethyltriazenes have been evaluated in the Ames test. The octanol-water partition coefficients (P) for these triazenes have been measured, and their electron distributions and molecular orbital energies were calculated using the MNDO semiempirical molecular orbital method. Molecular structures of three triazenes have been determined using X-ray crystallography. The mutagenicities of these five triazenes, which range from nearly inactive to very highly mutagenic, are well predicted by quantitative structure-activity relationships that had been derived previously for the mutagenicity of aryltriazenes. The form of these equations indicates that more hydrophobic and more electron-rich triazenes are more active in the Ames test. This supports the hypothesis that the ease of initial triazene activation by cytochrome P-450 governs the mutagenicity of these compounds.

Quantitative structure-activity relationships have been formulated for four sets of amidine inhibitors of trypsin. The quantitative results from these equations are compared with qualitative models constructed from the X-ray crystallographic coordinates of a benzamidine bound to trypsin. The good agreement between the mathematical and graphics models provides further support for the use of substituent constants and regression analysis in the study of enzyme-ligand interactions.

A quantitative structure-activity relationship (QSAR) (log K = 1.55 alpha + 0.64 log P - 2.07I1 - 3.28I2 + 6.94) has been formulated for the binding of a set of substituted benzenesulfonamides to human carbonic anhydrase. The binding constant (K) are from the studies of King and Burgen [Proc. R. Soc. Lond. B. 193:107-125 (1976)], sigma is the Hammett electronic substituent constant, P is the octanol/water partition coefficient, and I1 and I2 are indicator variables for meta and ortho substituents, respectively. The negative coefficients with the indicator variables suggest steric hindrance by these substituents in contrast to para substituents. Qualitative features of the QSAR are correlated with a color stereomolecular graphics model of the enzyme-inhibitor complex which was constructed from the X-ray crystallographic coordinates of the enzyme.

Phenol sulfotransferase (PST) catalyzes the sulfate conjugation of many phenolic drugs. Human liver contains thermostable (TS) and thermolabile forms of PST. Ion exchange chromatography shows that two isozymes of TS PST (peaks I and II) are present in human liver preparations. Rat liver contains four forms of PST that can be separated by ion exchange chromatography. Quantitative structure-activity relationship (QSAR) analysis was used to study phenolic substrates for both human and rat liver PST. Thirty-six substituted phenols were tested as substrates for partially purified human liver TS PST peak I. QSAR analysis resulted in derivation of the following equation: log 1/Km = 0.92 (+/- 0.18)log P - 1.48 (+/- 0.38)MR'4 - 0.64 (+/- 0.41)MR3 + 1.04 (+/- 0.63)MR2 + 0.67(+/- 0.44) sigma- + 4.03 (+/- 0.42). In this equation Km is the Michaelis constant, P is the octanol-water partition coefficient, MR is the molar refractivity of substituents at the 2-, 3-, and 4-positions, and sigma- is the Hammett constant. Values of log 1/Km calculated with this equation were highly correlated with log 1/Km values (r = 0.950) that were observed experimentally. Nine phenols were also tested as substrates for partially purified human liver TS PST peak II. Log 1/Km values for these compounds were significantly correlated for the two isozymes of TS PST (r = 0.992, p less than 0.001). QSAR analysis was also used to derive equations that described the behavior of phenolic substrates for rat liver PST forms I and II. These equations differed substantially from the equation derived for compounds tested with human liver TS PST peak I. Therefore, the characteristics of the active sites of human liver TS PST peak I and rat liver PST forms I and II appear to differ. Application of these equations may make it possible to predict Km values of phenolic substrates for human liver TS PST and for rat liver PST forms I and II.

The hydrolysis of a set of 28 X-phenyl hippurates by chymotrypsin was investigated. From the derived Km and kcat values a quantitative structure-activity relationship was developed. This equation shows that para substituents correlated by sigma- display only an electronic effect on the formation of the ES complex whereas meta hydrophobic substituents show a hydrophobic interaction correlated by pi in addition to their electronic effect. Meta polar substituents avoid contact with the enzyme and show only electronic effects on Km. Using the x-ray crystallographic coordinates for chymotrypsin and computer graphics, a model was constructed which is used to interpret the quantitative structure-activity relationship. As with a number of previously reported examples, we have found that when polar substituents have the option of binding to hydrophobic space or remaining in the aqueous phase they follow the latter possibility.