1262
Chem. Res. Toxicol. 1999, 12, 1262-1267
Structure-Toxicity Relationships for Benzenes Evaluated with Tetrahymena pyriformis T. Wayne Schultz* College of Veterinary Medicine, The University of Tennessee, P.O. Box 1071, Knoxville, Tennessee 37901-1071 Received April 27, 1999
Toxicity data for 200 substituted benzenes tested in the two-day Tetrahymena pyriformis population growth impairment assay representing the neutral narcosis, polar narcosis, respiratory uncoupling, and weak and strong electrophilic mechanisms of toxic action were evaluated. A quantitative structure-toxicity model correlating toxic potency [log(IGC50-1)] with hydrophobicity quantified by the 1-octanol/water partition coefficient (log Kow) and electrophilic reactivity quantified by the molecular orbital parameter, maximum superdelocalizability (Smax), was developed. This model [log(IGC50-1) ) 0.50(log Kow) + 9.85(Smax) - 3.47; n ) 197, r2 ) 0.816, s ) 0.34, F ) 429, Pr > F ) 0.0001] allows for the prediction of acute potency without the a priori identification of the mechanism of action. The examination of residuals reveals that neutral narcotics with high volatility (e.g., methyl- and chloro-substituted benzenes) and highly reactive fluoro- and nitro-containing derivatives are fitted poorly. A comparison of observed (obs) and predicted (pred) toxicities on the additional set of derivatives [log(obs IGC50-1) ) 1.05[log(pred IGC50-1)] + 0.02; n ) 20, r2 ) 0.979, s ) 0.13, F ) 825, Pr > F ) 0.0001] validated the model as a good predictor of toxicity regardless of the mechanism of toxic action.
Introduction Among the most prevalent industrial organic chemicals are substituted benzenes. The parent compound, benzene, has an equal number of carbon and hydrogen atoms. Benzene generally is envisioned as being a regular hexagon with bond angles of 120°, sp2-hybrid orbitals, and a fourth valence of π-bonds from p-orbitals extending equally around the ring. These characteristics impart the aromatic nature of the substance. With this delocalization, benzene does not exhibit the high reactivity typical of polyene compounds. However, this fact changes dramatically when benzene is substituted with unsaturated (i.e., π-bond-containing) substituents, especially in conjunction with leaving groups (1). Among the protozoan test systems, members of the Ciliophora, especially primitive free-living holotrichians, are frequently utilized in aquatic toxicity assessments. With such ciliates, it is possible to monitor a number of toxicological end points (2). Among these end points, the two most common are mortality and population growth impairment. The most extensive ciliate toxicity database is that for the TETRATOX assay, population growth inhibition in Tetrahymena pyriformis (2). While toxicity databases on a number of fish species, including the guppy (Poecilia reticulata) and zebrafish (Brachydanio rerio), exist, the largest and most chemically diverse of the fish data sets is that for 96-h flowthrough fathead minnow (Pimephales promelas) 50% mortality (3). Previous studies (2, 4, 5) have shown good agreement between fish toxicity and toxic potency measures in T. pyriformis. * To whom correspondence should be addressed. Phone: (865) 9745826. Fax: (865) 974-2215. E-mail:
[email protected].
Quantitative structure-activity relationships are powerful tools in predictive toxicology (6, 7). A simplistic toxicokinetic- and toxicodynamic-based approach to potency modeling was outlined by McFarland (8). He expressed chemical toxicity as a combination of uptake into or through biological membranes and the interaction of the toxicant with the site of action. McFarland’s approach is represented mathematically as
log(toxicity)-1 ) A(log of uptake) + B(log of interaction) + C Uptake of most industrial organic chemicals to the site of action is by passive diffusion and is best modeled by hydrophobicity, most often quantitated by the 1-octanol/ water partition coefficient (log Kow).1 Interaction of the chemical with the site of action is more complicated and is quantified by a number of molecular parameters that describe electronic and/or steric properties. The most descriptive interactive parameters appear to be quantum chemical parameters, especially ones based on frontier orbitals. Historically, the accuracy in predicting potency hinges in part on selecting the correct toxic mechanism (9). Since toxicity depends on a variety of physical and/or chemical interactions (10) between the substance and an frequently poorly defined molecular site of action, selection of the correct toxic mechanism is not an easy undertaking (7). Thus, models that accurately predict acute toxicity without first identifying toxic mechanisms are highly 1 Abbreviations: QSARs, quantitative structure-activity relationships; IGC50, 50% growth inhibitory concentration; log Kow, 1-octanol/ water partition coefficient; Smax, maximum superdelocalizability; obs, observed; pred, predicted.
10.1021/tx9900730 CCC: $18.00 © 1999 American Chemical Society Published on Web 11/19/1999
QSARs for Benzenes
desirable. Analyses in refs 11-13 suggest that a simple response surface developed from multiple regression analysis (14) could provide such a model. The aim of this investigation reported herein was the development and validation of a two-parameter structuretoxicity model of acute toxicity from data for a heterogeneous set of substituted benzenes representing a variety of known and unknown toxic mechanisms.
Materials and Methods Chemicals. Two hundred twenty benzene homologues representing several mechanisms of toxic action (5) were assessed. Caution: The following chemicals are hazardous and should be handled carefully. As reported here, several of these chemicals exhibit significant acute toxicity and are potential mutagens and skin sensitizers (1). Derivatives were selected a priori to ensure uniform distribution over the maximum range of values for both the hydrophobic and electrophilic descriptors. The homologues were obtained commercially (Aldrich Chemical Co., Milwaukee, WI; MTM Research Chemicals or Lancaster Synthesis Inc., Windham, NH) at sufficient purity (>95%) that made further purification unnecessary. Biological Data. Population growth impairment testing with the common ciliate T. pyriformis (strain GL-C) was conducted following the protocol described by Schultz (2). This 40-h assay is static in design and uses population density quantitated spectrophotometrically at 540 nm as its end point. The test protocol allows for eight to nine cell cycles in controls. Following range finding, each chemical was tested in three replicate tests (or assays). Two controls were used to provide a measure of the acceptability of the test by indicating the suitability of the medium and test conditions as well as a basis for interpreting data from other treatments. The first control had no test material and was inoculated with T. pyriformis. The other, a blank, had neither test material nor ciliates. Each test replicate consisted of six to eight different concentrations of each test material with duplicate flasks with each concentration. Only replicates with control absorbency values of >0.6 but F ) 0.0001. Nine derivatives were observed to be statistical outliers to eq 1. These compounds and their residual values are reported in Table 2. An examination of residual values showed them in general to have a random distribution (data not shown). However, typically nonreactive compounds (i.e., Smax ) 24) with a propensity to volatilize (i.e., low vapor pressures) were less toxic than predicted by eq 1, while highly reactive chemicals (i.e., Smax < 36) were more toxic than predicted by eq 1. Despite the observance of statistical outliers, no effort was made to improve the fit of the model by removal of homologues. Similarly, no effort was made to improve the model by the addition of other descriptors. In an effort to validate eq 1, a series of 20 additional benzene derivatives were tested and the observed toxicities compared with that predicted by eq 1. These additional homologues were identified prior to toxicity model development. Selection was based on their descriptor values. Data for these chemicals are reported in Table 3. On a linear scale, reactivity varied uniformly from 24 to 37. While on a log scale, hydrophobicity varied from -0.55 to 4.12; the vast majority of the values were between 2.0 and 4.0. The reduction in ranges for both parameters reflects the limited number of benzenes that are commercially available, having either high log Kow or high Smax values. Regression analysis of observed versus predicted toxicity yielded
log (obs IGC50-1) ) 1.05[log(pred IGC50-1)] + 0.02 (2) where n ) 20, r2 ) 0.979, s ) 0.13, F ) 825, and Pr > F ) 0.0001.
1264
Chem. Res. Toxicol., Vol. 12, No. 12, 1999
Schultz
Table 1. Toxicity to T. pyriformis [log(IGC50-1)], Octanol/Water Partitioning (log Kow), and Maximum Superdelocalizability (Smax) Valuesa compound
CAS number
benzyl alcohol benzylamine sec-phenethyl alcohol 3-phenyl-1-propanol 1-phenyl-2-butanol methoxybenzene 3-phenyl-1-butanol benzene 4-ethylbenzyl alcohol (()-2-phenyl-2-butanol 4-phenyl-1-butanol R,R-dimethylbenzenepropanol 1-phenyl-1-butanol ethoxybenzene methylbenzene thioanisole 5-phenyl-1-pentanol 4-methylanisole 1,1-diphenyl-2-propanol 4-biphenylmethanol 1,4-dimethylbenzene benzophenone (()-1,2-diphenyl-2-propanol 6-phenyl-1-hexanol isopropylbenzene biphenyl n-butylbenzene n-amylbenzene 4-ethylbiphenyl aniline 4-methylaniline 3-methylaniline 2-methylaniline 2-ethylaniline 2-phenyl-3-butyn-2-ol 3-ethylaniline 4-ethylaniline 4-butoxyaniline 2,6-diethylaniline 4-pentyloxyaniline 2,6-diisopropylaniline 4-butylaniline 4-methoxyphenol benzaldehyde phenol 3-methoxyphenol 3-ethoxy-4-hydroxybenzaldehyde acetophenone 4-methylphenol 2-methylphenol 3-methylphenol propiophenone 2,4-dimethylphenol 2-ethylphenol 4-ethylphenol 3-ethylphenol 2-allylphenol 2,3,6-trimethylphenol 2,4,6-trimethylphenol butyrophenone 3,4,5-trimethylphenol 2,3,5-trimethylphenol 4-isopropylbenzaldehyde valerophenone 4-propylphenol iodobenzene 4-tert-butylphenol 4-hexyloxyaniline 4-tert-pentylphenol 4-pentyloxybenzaldehyde heptanophenone octanophenone nonylphenol 1,3-dihydroxybenzene thiobenzamide
100-51-6 100-46-9 98-85-1 122-97-4 120055-09-6 100-66-3 2722-36-3 71-43-2 768-59-2 1565-75-9 3360-41-6 103-05-9 22144-60-1 103-73-1 108-88-3 100-68-5 10521-91-2 104-93-8 29338-49-6 3597-91-9 106-42-3 119-61-9 5342-87-0 2430-16-2 98-82-8 92-52-4 104-51-8 538-68-1 5707-44-8 62-53-3 106-49-0 108-44-1 95-53-4 578-54-1 127-66-2 587-02-0 589-16-2 4344-55-2 579-66-8 39905-50-5 24544-04-5 104-13-2 150-76-5 100-52-7 108-95-2 150-19-6 121-32-4 98-86-2 106-44-5 95-48-7 108-39-4 93-55-0 105-67-9 90-00-6 123-07-9 620-17-7 1745-81-9 2416-94-6 527-60-6 495-40-9 527-54-8 697-82-5 122-03-2 1009-14-9 645-56-7 591-50-4 98-54-4 39905-57-2 80-46-6 5736-91-4 1671-75-6 1674-37-9 104-40-5 108-46-3 2227-79-4
log Smax (IGC50-1) 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.27 0.27
-0.83 -0.24 -0.66 -0.21 -0.16 -0.10 0.01 -0.12 0.07 0.06 0.12 -0.07 -0.01 0.10 0.25 0.18 0.42 0.25 0.75 0.92 0.25 0.87 0.80 0.87 0.69 1.05 1.25 1.79 1.97 -0.23 -0.05 -0.28 -0.16 -0.22 -0.18 -0.03 0.03 0.61 0.31 0.97 0.76 1.07 -0.14 -0.20 -0.35 -0.33 0.02 -0.05 -0.16 -0.29 -0.08 0.05 0.14 0.16 0.21 0.29 0.33 0.28 0.42 0.30 0.93 0.36 0.67 0.56 0.64 0.50 0.91 1.38 1.23 1.18 1.56 1.89 2.47 -0.65 0.09
log Kow
compound
1.05m 1.09m 1.42m 1.88m 2.02e 2.11m 2.11e 2.13m 2.13e 2.34e 2.35m 2.42e 2.47e 2.51m 2.73m 2.74m 2.77e 2.81m 2.93e 2.99e 3.15m 3.18m 3.23e 3.30e 3.66m 3.98m 4.26m 4.90m 5.06e 0.90m 1.39m 1.40m 1.43m 1.74m 1.88e 1.94e 1.96m 2.59e 2.87e 3.12e 3.18m 3.18m 1.34m 1.48m 1.50m 1.58m 1.58m 1.63m 1.97m 1.98m 1.98m 2.19m 2.35m 2.47m 2.50m 2.50m 2.55e 2.67m 2.73m 2.77m 2.87e 2.92e 2.92e 3.17e 3.20m 3.25m 3.31m 3.65e 3.83e 3.89e 4.23e 4.75e 5.76m 0.80m 1.50m
4-chlorobenzamide 4-chloroaniline 2-chloroaniline 3-chloroaniline 2-tolunitrile benzyl chloride 2-chloro-4-methylaniline 4-chloroanisole olivetol 4-hexylresorcinol salicylaldehyde 2-hydroxy-4-methoxyacetophenone 4-chlorophenol chlorobenzene phenethyl bromide 4-chloro-3-methylphenol 4-chlorobenzophenone 2-amino-5-chlorobenzonitrile 5-bromovanillin 3-chlorophenol 3-chloro-5-methoxyphenol 3,4-dichloroaniline 2,5-dichloroaniline 3,5-dichloroaniine bromobenzene 2,4-dichlorophenol 4-chloro-3,5-dimethylphenol 4-bromotoluene 3,4-dichlorotoluene 1-bromo-4-ethylbenzene 3-nitroaniline nitrobenzene 3-nitroanisole 2-nitrotoluene 4-nitrotoluene 4-ethoxy-2-nitroaniline 3-nitrotoluene R,R,R-4-tetrafluoro-m-toluidine 4-nitrophenetole 1,2-dimethyl-3-nitrobenzene 1,2-dimethyl-4-nitrobenzene 1,2-dichlorobenzene 3,5-dichlorophenol 2,4,5-trichloroaniline 1,4-dibromobenzene 5-hydroxy-2-nitrobenzaldehyde 2-nitrophenol 2-nitroaniline 2-chloro-4-nitroaniline 4-chlorobenzaldehyde pentafluorobenzaldehyde 2,4,6-trichlorophenol 2,4,5-trichlorophenol 1,2,4-trichlorobenzene 2,4,6-tribromophenol 1,3,5-trichlorobenzene pentafluoroaniline 1-fluoro-4-nitrobenzene 1-chloro-3-nitrobenzene 4-chloro-2-nitrotoluene 2-chloro-6-nitrotoluene 2,3,5,6-tetrachloroaniline 2,3,4,5-tetrachloroaniline 1,2,4,5-tetrachlorobenzene 4,5-dichloro-2-nitroaniline phenyl isothiocyanate 2,4-dichloro-6-nitroaniline 3,5-dibromosalicylaldehyde 2,3,4,6-tetrachlorophenol 2,3,4,5-tetrachlorophenol pentachloronitrobenzene pentachlorobenzene pentachloroanisole 2-chloro-5-nitrobenzaldehyde
CAS number
log Smax (IGC50-1)
log Kow
104-86-9 106-47-8 95-51-2 108-42-9 529-19-1 100-44-7 615-65-6 623-12-1 500-66-3 136-77-6 90-02-8 552-41-0
0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.28 0.28
0.16 0.05 -0.17 0.22 -0.24 0.10 0.18 0.60 1.31 1.80 0.42 0.55
1.81e 1.83m 1.88m 1.88m 2.21m 2.30m 2.41e 2.79m 3.42e 3.45m 1.81m 1.98m
106-48-9 108-90-7 103-63-9 59-50-7 134-85-0 5922-60-1 2973-76-4 108-43-0 65262-96-6 554-00-7 95-82-9 626-43-7 108-86-1 120-83-2 88-04-0 106-38-7 95-75-0 1585-07-5 99-09-2 98-95-3 555-03-3 88-72-2 99-99-0 616-86-4 99-08-1 2357-47-3 100-29-8 83-41-0 99-51-4 95-50-1 591-35-5 636-30-6 106-37-6 42454-06-8 88-75-5 88-74-4 121-87-9 104-88-1 653-37-2 88-06-2 95-95-4 120-82-1 118-79-6 108-70-3 771-60-8 350-46-9 121-73-3 89-59-8 83-42-1 3481-20-7 634-83-3 95-94-3 6641-64-1 103-72-0 2683-43-4 90-59-5 58-90-2 4901-51-3 82-68-8 608-93-5 1825-21-4 6361-21-3
0.28 0.28 0.28 0.28 0.28 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.29 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.30 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.32 0.32 0.32 0.32 0.32 0.32 0.32 0.32 0.33 0.33 0.33 0.33 0.33 0.33 0.33 0.33 0.33 0.34
0.54 -0.13 0.50 0.80 1.50 0.44 0.62 0.87 0.76 0.56 0.58 0.71 0.75 1.04 1.20 1.00 1.07 1.10 0.03 0.14 0.72 0.26 0.65 0.76 0.42 0.77 0.83 0.56 0.59 1.00 1.56 1.30 0.68 0.33 0.67 0.08 0.75 0.40 0.82 1.41 2.10 1.10 1.91 0.87 0.26 0.10 0.73 0.82 0.68 1.76 1.96 2.00 1.66 1.41 1.26 1.65 2.18 2.72 NTAS NTAS NTAS 0.53
2.39m 2.84m 3.09m 3.10m 3.97e 1.79e 1.92e 2.50m 2.50e 2.78m 2.75m 2.90m 2.99m 3.17m 3.48e 3.50e 3.95m 4.03e 1.43m 1.85m 2.17m 2.30m 2.37m 2.39e 2.45m 2.51e 2.53m 2.83m 2.91m 3.38m 3.61m 3.69m 3.79m 1.75e 1.77m 1.85m 2.05e 2.13m 2.39e 3.69m 3.72m 4.02m 4.08m 4.19m 1.87e 1.89m 2.47m 3.05m 3.09m 4.10m 4.27m 4.63m 3.21e 3.28m 3.33e 3.42e 3.88m 4.21m 4.64m 5.17m 5.45m 2.25e
QSARs for Benzenes
Chem. Res. Toxicol., Vol. 12, No. 12, 1999 1265 Table 1. Continued
compound 1-chloro-4-nitrobenzene 4-chloro-3-nitrophenol 1-bromo-2-nitrobenzene 1-chloro-2-nitrobenzene 1-bromo-3-nitrobenzene 3-(trifluoromethyl)-4-nitrophenol 1-fluoro-3-iodo-5-nitrobenzene 2-bromo-1-methyl-5-nitrobenzene 3,4,5,6-tetrabromo-2methylphenol pentachlorophenol 3-nitrobenzonitrile 1,3-dinitrobenzene 1,2-dinitrobenzene 1-methyl-2,4-dinitrobenzene 2,5-dichloro-1-nitrobenzene 2,3-dichloro-1-nitrobenzene 2,4-dichloro-1-nitrobenzene 3,5-dichloro-1-nitrobenzene 3,4-dichloro-1-nitrobenzene pentabromophenol 2,4-dinitroaniline 2,6-dinitroaniline 3-chloro-4-fluoro-1-nitrobenzene 3,4-dinitrobenzyl alcohol 1,4-dinitrobenzene a
Smax
log (IGC50-1)
log Kow
100-00-5 610-78-6 577-19-5 88-73-3 585-79-5 88-30-2 3819-88-3 7149-70-4 576-55-6
0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34 0.34
0.43 1.27 0.75 0.68 1.03 1.65 1.09 1.16 2.57
2.39m 2.46e 2.51m 2.52m 2.64m 2.77e 3.15e 3.25e 4.47e
87-86-5 619-24-9 99-65-0 528-29-0 121-14-2 89-61-2 3209-22-1 611-06-3 618-62-2 99-54-7 608-71-9 97-02-9 606-22-4 350-30-1 79544-31-3 100-25-4
0.34 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.36 0.36 0.36 0.37 0.37
2.07 0.45 0.76 1.25 0.87 1.13 1.07 0.99 1.13 1.16 2.66 0.72 0.84 0.80 1.09 1.30
5.18m 1.17m 1.49m 1.69m 1.98m 3.03m 3.05m 3.09m 3.09m 3.12m 4.85e 1.72e 1.79m 2.77e 0.59e 1.47m
CAS number
compound
CAS number
6-chloro-2,4-dinitroaniline 2-bromo-4,6-dinitroaniline 1,2,4-trichloro-5-nitrobenzene 1,2,3-trichloro-4-nitrobenzene 1,3,5-trichloro-2-nitrobenzene 2,6-dinitrophenol 2,4-dinitrophenol 2,5-dinitrophenol 4,6-dinitro-2-methylphenol 4-tert-butyl-2,6-dinitrophenol 1,2,3-trifluoro-4-nitrobenzene 2,3,4,5-tetrachloronitrobenzene 2,3,5,6-tetrachloronitrobenzene 1,5-difluoro-2,4-dinitrobenzene 2,3,4,6-tetrafluoronitrobenzene 1-iodo-2,4-dinitrobenzene 1-fluoro-2,4-dinitrobenzene pentafluoronitrobenzene 1-bromo-2,4-dinitrobenzene 1,2-dichloro-4,5-dinitrobenzene 1,5-dichloro-2,3-dinitrobenzene 1-chloro-2,4-dinitrobenzene 1,3,5-trichloro-2,4-dinitrobenzene 1,3-dinitro-2,4,5-trichlorobenzene 4-chloro-3,5-dinitrobenzonitrile 1,4-dinitrotetrachlorobenzene
3531-19-9 1817-73-8 89-69-0 17700-09-3 18708-70-8 573-56-8 51-28-5 329-71-5 534-52-1 4097-49-8 771-69-7 879-39-0 117-18-0 327-92-4 314-41-0 709-49-9 70-34-8 880-78-4 584-48-5 6306-39-4 28689-08-9 97-00-7 6284-83-9 2678-21-9 1930-72-9 20098-38-8
log Smax (IGC50-1) 0.37 0.37 0.37 0.37 0.37 0.38 0.38 0.38 0.38 0.38 0.39 0.39 0.39 0.40 0.40 0.40 0.41 0.41 0.42 0.42 0.42 0.43 0.44 0.45 0.47 0.47
1.12 1.24 1.53 1.51 1.43 0.83 1.06 1.04 1.73 1.80 1.89 1.78 1.82 2.08 1.87 2.12 1.71 2.43 2.31 2.21 2.42 2.16 2.19 2.60 2.66 2.82
log Kow 2.46e 2.61e 3.47m 3.61m 3.69m 1.33m 1.54m 1.86m 2.12m 3.61e 2.01e 3.93m 4.38m 1.31e 1.86e 2.50e 1.47e 2.00e 2.29e 2.43e 2.85e 2.14e 2.97e 3.05e 1.37e 3.44e
e designates an estimated value. m designates a measured value. NTAS, not toxic at saturation.
be nonbinding to macromolecules and exhibit acute toxicities that are directly related to log Kow. These chemicals cause mostly reversible physiological alterations. Chemicals acting as neutral narcotics are deemed electrophilically unreactive. Their toxicity is proportional to their concentration at the site of action and is merely due to the extent of membrane perturbation (20). Because narcosis depends only on the uptake of chemicals, the modeling of toxic potency of narcotics yields simple log Kow-dependent models. In contrast to neutral narcotics, bioreactive chemicals are chemicals that have a positive electronic and/or steric interaction with a biological system. Such interactions may be direct or mediated by metabolism. Moreover, bioreactivity may be further subdivided into noncovalent and covalent mechanisms. Noncovalent mechanisms react reversibly. Covalent mechanisms are irreversible. Noncovalent-mediated bioreactive mechanisms include polar narcosis (21, 22) and weak acid respiratory uncou-
Table 2. Statistical Outliers and Residual Values from eq 1 compound
residual value
chlorobenzene 1,3,5-trichlorobenzene 2,3,5,6-tetrachloronitrobenzene 1,4-dibromobenzene 1,5-difluoro-2,4-dinitrobenzene 4-hexylresorcinol pentafluoronitrobenzene 2,3,4,5-tetrachlorophenol 4-chloro-3,5-dinitrobenzonitrile
-0.84 -0.81 -0.74 -0.70 0.95 0.89 0.86 0.84 0.82
The statistics for eq 2, slope of ∼1 and an intercept of ∼0, demonstrate a nearly perfect 1:1 correlation between observed and predicted potencies.
Discussion Most industrial organic chemicals exhibit a narcosis mode of toxic action (19). Such chemicals are thought to
Table 3. Descriptor and Toxicity Values for the Validation Seta compound 4-ethylbenzyl alcohol 3-aminobenzyl alcohol 3,4-dimethylaniline 4-isopropylaniline 4-hydroxyphenethyl alcohol 4-hydroxypropiophenone 4-tert-pentylphenol 2-(4-chlorophenyl)ethylamine vanillin 4-iodophenol 4-bromobenzophenone 4-methyl-2-nitroaniline 4-bromo-6-chloro-2-methylphenol 2,4-dibromophenol 4-bromo-2,6-dichlorophenol pentafluorophenol 4-nitrobenzyl chloride 4-chloro-6-nitro-3-methylphenol 2,6-diiodo-4-nitrophenol 2,4-dichloro-6-nitrophenol a
CAS number
Smax
log Kow
log(pred IGC50-1)
log(obs IGC50-1)
768-59-2 1877-77-6 95-64-7 99-88-7 501-94-0 70-70-2 80-46-6 156-41-2 121-33-5 540-38-5 90-90-4 89-62-3 7530-27-0 615-58-7 697-86-9 771-61-9 100-14-1 7147-89-9 305-85-1 609-89-2
0.24 0.25 0.25 0.25 0.26 0.26 0.26 0.27 0.28 0.29 0.29 0.30 0.30 0.31 0.32 0.33 0.34 0.34 0.36 0.37
2.13e -0.55m 1.86e 2.47m 0.52e 2.03m 3.83e 2.00e 1.21m 2.90m 4.12e 1.82m 3.61e 3.25m 3.52e 3.23m 2.45e 2.93m 3.52e 3.07e
-0.04 -1.28 -0.08 0.23 -0.65 0.11 1.01 0.19 -0.11 0.84 1.34 0.40 1.29 1.21 1.44 1.40 1.10 1.34 1.84 1.71
0.07 -1.13 -0.16 0.22 -0.83 0.05 1.23 0.14 -0.03 0.85 1.26 0.37 1.28 1.40 1.78 1.63 1.18 1.63 1.81 1.75
e designates an estimated value. m designates a measured value.
1266
Chem. Res. Toxicol., Vol. 12, No. 12, 1999
pling (21, 23). Polar narcotics represent the majority of aromatic compounds with strong electron-releasing amino or hydroxy moieties (12). As with nonpolar narcotics, these effects were reversible. Polar narcotics (24) exhibit effects similar to neutral polar narcotics, but at potency levels greater than estimated by their hydrophobicity and the neutral narcosis model (21). Weak acid respiratory uncouplers disrupt ATP synthesis and are generally bulky and electronegative. They cause the inner mitochondrial membrane to be permeable to hydrogen ions, thereby disrupting the hydrogen ion gradient (25, 26). Nonreversible, bioreactivity results in chemical changes in biological systems (20). The term nonreversible bioreactive is not specific to a certain reaction; instead, it includes a number of competing processes and different chemical reactivity mechanisms. Most of these mechanisms involve alkylation or arylation to soft nucleophiles, in particular thiol and amino moieties associated with proteins. Such chemical reactions include Schiff-base formation, Michael and Michael-type acceptance, and nucleophilic substitution (27). Recent developments in toxicity-based relationships include the mechanism of action approach (9). However, problems associated with assignment of an a priori mechanism of action have deterred efforts to use this approach for prediction of toxic potency of complex chemicals. As noted by Schultz et al. (5), while recent structure-toxicity studies of industrial narcotics and soft electrophiles point toward a mechanism-based approach, due to available data these studies tend to be more classbased investigations. The examination of Smax values among chemicals associated with a particular mechanism reveals trends. Classic industrial aromatic neutral organic compounds such as alkyl-substituted benzenes exhibit Smax values of 24. Polar narcotics (e.g., most anilines and phenols) exhibit Smax values of 25-26. Weak acid uncouplers (e.g., 2,4-dinitrophenol and pentachlorphenol) exhibit Smax values of 34-38. Strongly reactive skin-sensitizing agents (e.g., halo-substituted dinitrobenzenes) exhibit Smax values in the 40s. Since the addition of leaving groups (i.e., halogens) to any of the above groups increases the Smax value in proportion to the number and type of halogens present, there is considerable overlap in Smax values and mechanisms of toxic action. While the toxicity data on reactive electrophiles are not as robust as those for narcotics, one structural indicator of soft electrophilicity is having an unsaturated functional group attached to an aromatic carbon (e.g., nitrobenzene) especially when substituted with leaving groups (i.e., halogens) (1). This alone does not identify the mechanism of action. Thus, a method that would aid in predicting toxic potency without identification of mechanism of toxic action is advantageous. Two types of derivatives, catechols and benzoquinones, were purposefully not included in the data sets. Schultz et al. (17) noted that catechols are capable of undergoing tautomerization to more electrophilic semiquinones. Benzoquinones based on their specific structure and oneelectron reduction potential react via a series of mixed and competing mechanisms (18). Since both of these subclasses are small and can be easily identified from molecular structure, these deletions do not significantly detract from the model.
Schultz
Recent efforts (1, 12, 13) have modeled toxicity across chemical classes and molecular mechanisms. However, these studies have been on much more structurally limited groups of chemicals. Equation 1 presented herein strongly suggests that at least for benzenes, a twoparameter model, one for biouptake and the other for orbital-controlled electrophilicity, is a good method. This model allows aquatic toxicity of substituted benzenes representing the neutral narcosis, polar narcosis, respiratory uncoupling, weak electrophilic, and strong electrophilic mechanisms to be modeled concomitantly.
Acknowledgment. This investigation was supported in part by The University of Tennessee Center of Excellence in Livestock Disease and Human Health. Gratitude is expressed to Ms. Susan Bryant and Mr. Glendon Sinks for their assistance with data procurement.
References (1) Cronin, M. T. D., Gregory, B. W., and Schultz, T. W. (1998) Response surface-based analyses of nitrobenzene toxicity to Tetrahymena pyriformis. Chem. Res. Toxicol. 11, 902-908. (2) Schultz, T. W. (1997) TETRATOX: The Tetrahymena pyriformis population growth impairment endpoint- A surrogate for fish lethality. Toxicol. Methods 7, 289-309. (3) Russom, C. L., Bradbury, S. P., Broderius, S. J., Hammermeister, D. E., and Drummond, R. A. (1997) Predicting modes of toxic action from chemical structure: Acute toxicity in the fathead minnow (Pimephales promelas). Environ. Toxicol. Chem. 16, 948967. (4) Bearden, A. P., and Schultz T. W. (1998) Comparison of Tetrahymena and Pimephales toxicity based on mechanism of action. SAR QSAR Environ. Res. 9, 127-153. (5) Schultz, T. W., Sinks, G. D., and Bearden, A. P. (1998) QSAR in aquatic toxicology: A mechanism of action approach comparing toxic potency to Pimephales promelas, Tetrahymena pyriformis, and Vibrio fischeri. In Comparative QSAR (Devillers, J., Ed.) pp 52-109, Taylor and Francis. (6) Karcher, W., and Devillers, J. (1990) SAR and QSAR in environmental chemistry and toxicology: Scientific tool or wishful thinking? In Practical Applications of Quantitative StructureActivity Relationships (QSAR) in Environmental Chemistry and Toxicology (Karcher, W., and Devillers, J., Eds.) pp 1-12, Kluwer Academic, Dordrecht, The Netherlands. (7) Bradbury, S. P. (1994) Predicting modes of toxic action form chemical structure: An overview. SAR QSAR Environ. Res. 2, 89-104. (8) McFarland, J. W. (1970) On the parabolic relationship between drug potency and hydrophobicity. J. Med. Chem. 13, 1092-1196. (9) Cronin, M. T. D., and Dearden, J. C. (1995) Review: QSAR in toxicology. 1. Prediction of aquatic toxicity. Quant. Struct.-Act. Relat. 14, 1-7. (10) Nendza, M., and Russom, C. L. (1991) QSAR modeling of the ERL-D fathead minnow acute toxicity database. Xenobiotica 21, 147-170. (11) Veith, G. D., and Mekenyan, O. G. (1993) A QSAR approach for estimating the aquatic toxicity of soft electrophiles (QSAR for soft electrophiles). Quant. Struct.-Act. Relat. 12, 349-356. (12) Cronin, M. T. D., and Schultz, T. W. (1996) Structure-toxicity relationships for phenols to Tetrahymena pyriformis. Chemosphere 32, 1453-1468. (13) Karabunarliev, S., Mekenyan, O. G., Karcher, W., Russom, C. L., and Bradbury, S. P. (1996) Quantum-chemical descriptors for estimating the acute toxicity of substituted benzenes to the guppy (Poecilia reticulata) and the fathead minnow (Pimephales promelas). Quant. Struct.-Act. Relat. 15, 311-320. (14) Devillers, J., and Lipnick, R. L. (1990) Practical applications of regression analysis in environmental QSAR studies. In Practical Applications of Quantitative Structure-Activity Relationships (QSAR) in Environmental Chemistry and Toxicology (Karcher, W., and Devillers, J., Eds.) pp 129-144, Kluwer Academic, Dordrecht, The Netherlands. (15) SAS Institute Inc. (1989) SAS/STAT User’s Guide, version 6, 4th ed., Vol. 2, pp 846, Cary, NC. (16) Mekenyan, O. G., Bonchev, D., Trinajstic, N., and Peitchev, D. (1986) Modeling the interaction of small molecules with biomacromolecules. II. A generalized concept for biological interaction. Arzneim.-Forsch. 24, 421-424.
QSARs for Benzenes (17) Schultz, T. W., Sinks, G. D., and Cronin, M. T. D. (1997) Identification of mechanism of toxic action of phenols to Tetrahymena pyriformis from molecular structure. In Quantitative Structure-Activity Relationships in Environmental Sciences (Fredenslund, F. C., and Schu¨u¨rmann, G., Eds.) Vol. VII, pp 329-342, SETAC Press, Pensacola, FL. (18) Schultz, T. W., Sinks, G. D., and Cronin, M. T. D. (1997) Quinoneinduced aquatic toxicity to Tetrahymena: Structure-activity relationships. Aquat. Toxicol. 39, 267-278. (19) Bradbury, S. P., and Lipnick, R. L. (1990) Introduction: Structural properties for determining mechanisms of toxic action. Environ. Health Perspect. 87, 181-182. (20) Mekenyan, O. G., and Veith, G. D. (1994) The electronic factor in QSAR: MO-parameters, competing interactions, reactivity, and toxicity. SAR QSAR Environ. Res. 2, 129-143. (21) Schultz, T. W., Holcombe, G. W., and Phipps, G. L. (1986) Relationships of quantitative structure-activity of selected phenols in the Pimephales promelas and Tetrahymena pyriformis test systems. Ecotoxicol. Environ. Saf. 12, 146-153. (22) Veith, G. D., and Broderius, S. J. (1987) Structure-toxicity relationships for industrial chemicals causing type II narcosis
Chem. Res. Toxicol., Vol. 12, No. 12, 1999 1267
(23)
(24)
(25) (26) (27)
syndrome. In QSAR in Environmental Toxicology (Kaiser, K. L. E., Ed.) Vol. II, pp 385-391, D. Reidel, Dordrecht, The Netherlands. Cajina-Quezada, M., and Schultz, T. W. (1990) Structure-toxicity relationships for selected weak acid respiratory uncouplers. Aquat. Toxicol. 17, 239-252. Veith, G. D., and Broderius, S. J. (1990) Rules for distinguishing toxicants that cause type I and type II narcosis syndrome. Environ. Health Perspect. 87, 207-211. McLaughlin, S. G. A., and Dilger, J. P. (1980) Transport of protons across membranes by weak acids. Physiol. Rev. 60, 825-862. Terada, H. (1990) Uncouplers of oxidative phosphorylation. Environ. Health Perspect. 87, 213-218. Karabunarliev, S., Mekenyan, O. G., Karcher, W., Russom, C. L., and Bradbury, S. P. (1996) Quantum-chemical descriptors for estimating the acute toxicity of electrophiles to the fathead minnow (Pimephales promelas): An analysis based on molecular mechanisms. Quant. Struct.-Act. Relat. 15, 320-310.
TX9900730
