QSAR study for the toxicity of anilines and phenols to aquatic organisms

Yuan Xing, Lu Guanghua, Zhao Yuanhui
(Department of Environmental Science, Northeast Normal University, Changchun, 130024, China)

Received Sep. 17, 2000; Supported by the National Natural Scientific Foundation of China (No. 29877004)

Abstract96h-LC50 of substituted anilines and phenols to carp (Cyprinus carpio), 15min -EC50 to Photobacterium phosphoreum and 48h-EC50 to the algae (Scenedesmus obliquus) were determined, respectively. There is obvious correlation among the toxicity values of the three kinds of organisms. The energy of the lowest unoccupied molecular orbital (ELUMO) was calculated by the quantum chemical method MOPAC6.0-AM1. By using ELUMO and the hydrophobicity parameter logP the quantitative structure-activity relationship models (QSARs) were developed:
Carp: log1/LC50=3.300.78ELUMO+0.33logP, n=16, R=0.869, S.E.=0.335, F=43.22, P=0.000.
Ph. Phosphoreum: log1/EC50=2.460.51ELUMO+0.43logP, n=16, R=0.759, S.E.=0.441, F=18.85, P=0.000.
Algae: log1/EC50=2.520.59ELUMO+0.32logP, n=14, R=0.851, S.E.=0.326, F=28.60, P=0.000.
For those compounds containing polyhalogen or nitrito-group, their toxicity may be related chiefly to the intracellular reactivity and hydrophobicity, while for the other anilines and phenols, their toxicity is controlled mainly by hydrophobicity instead of electronic factor.

Most anilines and phenols constitute a class of industrial chemicals that are present in the Songhua River of Jilin Province, China, and probably in other industrialized countries, so it is beneficial to make a deep study of their potential hazard to aquatic organism. The carp (Cyprinus carpio) is major fish in China. Since the algae are the primary producers in many aquatic ecosystems, their susceptibility to the contaminant has been the subject of many reports. In view of these, we have measured 96-LC50 values of 16 anilines and phenols to carp and 48h-EC50 values to the algae (Scenedesmus obliquus). Furthermore, the toxicity to Photobacterium phosphoreum also has been determined in the Microtox assay, in order to further evaluate the usefulness of this assay for hazard assessment.

The one-year-old carps supplied by the Changchun Aquatic Institute were used in the experiment. The average length and weight were 11.6cm (卤2.3cm) and 23.8g (卤6.4g),respectively.The fish were reared under the laboratory condition for two weeks after they were sterilized in 5% (w/v) salt water. The death rate during the domestication was lower than 0.5%. The test water was dechlorinated tap water with 21.45mg/l chlorine. Temperature of the test water: 15-18掳; Dissolved oxygen: 6.35mg/l; and pH: 7.0-7.5. During the experiment, the direct sunlight was avoided thus lest photolysis of chemicals occur. Test aquaria were 60l glass tanks (L脳W脳H=50cm脳30cm脳40cm).Each contained 20l of test water and ten randomly selected fishes. Acetone was used as solvent with contents of 0.05-0.1% (v/v). During the 96h experiment, the test water was replaced twice a day and 10 l each time. The aquaria with same number of fishes and equal amount of solvent served as controls.
Scenedesmus obliquus was supplied by the Institute of Hydrobiology, Chinese Academy of Science. The method was referred to the Algae Inhibition Test (OECD, 1981). The culture was maintained in liquid medium (OJEC, 1988). The algae in logarithmic growing period were inoculated into 250 ml of Erlenmayer flasks, amounting to 60 ml of culture media, compound and the algae. The initial algae cell concentration in the test culture was approximately 1脳10 cells/ml. The culture was incubated under a continuous light at 20卤掳for 48 h and the average illumination intensity was about 4000LuX produced by white flourescent lamp. Growth was monitored by electron microscope (400 times). Data were handled according to the following formulas:=ln(N/N)/(t-t) , whereis the average specific growth rate, N is the initial cell concentration, N is the cell concentration after being cultured for 48h, and t-t is experimental period(48h in this work). I=[(b)-(tOX)]/(b)脳100%, where I is the inhibition rate,(b) is the average specific growth rate of the control,(tOX) is the average specific growth rate of added toxic compound.
The tests with Photobacterium phosphoreum were conducted using the Microtox toxicity analyzer (DXY-2, made by the Institute of Soil Science, Academia Sinica, Nanjing, China). The concentration values causing 50% reduction of bioluminescence after 15 min of exposure (15-min EC50) were performed at 20掳according to the procedures described in the Instrumental Manual.
For each compound at least five concentration gradients were planned, and there were two replicates at each concentration and control.

The energy values of the lowest unoccupied molecular orbital (ELUMO) of 20 compounds were calculated by the quantum chemical method MOPAC6.0-AM1 on energy-minimized structures. This method can automatically optimize the bond length, the bond angle and the twist angle, and yield a lot of structure information. The method can calculate those compounds whose atom numbers are smaller than 40.

able 1. The toxicity values to the aquatic organisms and phisico-chemical parameter values of substituted anilines and phenols

Compounds log1/LC50
Exp.    Pre.
Exp.   Pre.
Exp.   Pre.
Aniline 2.95  3.13 2.64  2.55 2.56    2.43 0.64 1.03
3-Chloroaniline           3.79 2.85  3.17 2.79    2.99 0.16 1.88
2-Chloroaniline 3.78  3.72 2.96  3.10 2.89  2.94 0.19 1.76
2,3-Dichloroaniline 4.37  4.25 4.51    3.59 3.98  3.39 -0.20 2.44
2,4-Dichloroaniline 4.65  4.34 4.06    3.73 3.74  3.48 -0.17 2.80*
2,5-Dichloroaniline 4.43  4.41 3.96    3.78 3.82  3.54 -0.26 2.80*
2,4,6-Tribromoaniline 4.09  4.88 4.76    4.35 4.37  3.98 -0.34 4.03
3-Bromoaniline           3.83 3.04  3.25 2.80  3.05 0.19 2.10
4-Bromoaniline 3.71  3.72 3.47    3.21            2.97 0.40 2.26
2-Methylaniline            3.29            2.74 2.34  2.58 0.59 1.40
3-Methylaniline 3.06  3.29            2.74            2.58 0.60 1.41
4-Methylaniline 2.91  3.28            2.73            2.57 0.61 1.39
Phenol 3.70  3.46 2.63  2.86 2.46  2.72 0.40 1.46
2-Methylphenol 3.54  3.67 2.44  3.14            2.92 0.41 2.12
3-Methylphenol 4.14  3.65           3.10            2.89 0.38 1.98
4-Methylphenol 3.85  3.59 3.16    3.05            2.85 0.43 1.93
2,4-Dichlorophenol 4.77  4.59 4.02   3.92 3.62  3.68 -0.43 2.92*
2,4,6-Trichlorophenol 5.14 3.85   4.46 3.81  4.15 -0.82 3.69
Pentachlorophenol 6.29  6.06 5.11   5.35 4.63  4.95 -1.43 5.04
2,4-Dinitrophenol 5.02  5.21 4.09   4.04 4.16  4.08 -1.81 1.54

a. Exp. is the experimental log1/LC50 (mol/l) to fish and Pre is the predicted values calculated from equation (1);
b. Exp. is the experimental log1/EC50 (mol/l) to Ph. phosphoreum and Pre is the predicted values calculated from equation (2);
c. Exp. is the experimental log1/EC50 (mol/l) to algae and Pre is the predicted values calculated from equation (3).
d. The logP values identified with an asterisk are calculated using a fragment constant method (Lyman, 1982), the rest are measured values obtained from literature (Deneer, 1987, Veith, 1993).

The toxicological data and parameters of the tested compounds are listed in Table 1. There are obvious correlations among the toxicity values of the three kinds of aquatic organisms. The correlation coefficient R is up to 0.80, 0.81 and 0.98 between carp and Ph. Phosphoreum, carp and algae, and Ph. Phosphoreum and algae, respectively.
Veith (1993) established a set of QSARs for aromatic chemicals by respectively using average superdelocalizability (image), ELUMO and logP to explain the variation of acute toxicity of substituted benzenes, phenols, and anilines to fish. The square of correlation coefficient R is up to 0.81 identically by using either image and logP or ELUMO and logP. Both image and ELUMO show the tendency of chemicals to undergo orbital-controlled reactions, and R is 0.94 between them. The QSAR for acute toxicity using these molecular descriptors defined a toxicity plane, which included several modes of toxic action. Type () narcotics are chemicals located in the region of low reactivity where toxicity varies with hodrophobicity alone. Type (I) narcotics are more toxic than Type () narcotics at similar values of logP, and the increase can be explained by stronger electronic interactions with cellular soft necleophiles. Dearden (1995) analyzed the data on acute toxicity of nitrobenzenes to Tetrahymena pyriformis and concluded that the toxicity of 2- and 3- substituted nitrobenzenes is probably controlled largely by hydrophobic and electronic factors.
On the basis of the above work, logP and ELUMO are selected as the parameters to establish the QSARs in this paper. By the multiple linear regression analyses of the toxicity values and parameters listed in Table 1, a series of QSAR equations are developed as follows:
Carp: log1/LC50=3.300.78ELUMO+0.33logP, n=16, R=0.869, S.E.=0.335, F=43.22, P=0.000 (1).
Ph. Phosphoreum: log1/EC50=2.460.51ELUMO+0.43logP, n=16, R=0.759, S.E.=0.441, F=18.85, P=0.000 (2).
Algae: log1/EC50=2.520.59ELUMO+0.32logP, n=14, R=0.851, S.E.=0.326, F=28.60, P=0.000 (3).
Where n is the number of compounds; R is the square of correlation coefficient; S.E. is the standard error; F is the mean square radio and P is the significant level.
The regressions of all equations are very obvious. logP is the logarithm of hydrophobicity parameter; the higher the logP values, the stronger the hydrophobicity, the easier the compound is bioconcentrated in a organism; ELUMO is an electrophilicity parameter, and it appears in directly proportional to the electronic affinity of the compound. The lower the ELUMO values, the stronger the electrophilicity (Wang, 1981). Both the data in Table 1 and above QSAR equations show that the toxicity of anilines and phenols to the three kinds of aquatic organisms is related chiefly to their ability to the penetrate into the cell by to the electronic interactions of the chemicals with the active site of action through a variety of electronic processes. To those compounds containing one -NO or polyhalogen, their logP values are higher and ELUMO values are negative, such as 2,4-Dinitrophenol and dichloroanilines, their toxicity is probably controlled not only by hydrophobic but also by electronic factors. This result is consistent with Veith’s (1993) and Dearden’s (1995). To those compounds such as aniline, methylanilines and phenol, their ELUMO values are positive and the contribution of ELUMO to the toxicity is negative. Therefore their toxicity is controlled mainly by hydrophobicity instead of electronic factor.
The equation (1), (2) and (3) are used to estimate the toxicity of 20 compounds listed in Table 1, and most compounds fit well. In this study, the compounds contain varied groups, such as -OH, -NH -CH, -NO, -Cl and -Br et al., so this QSAR equation for anilines and phenols can safely be used for predictive purposes.