Toxicity QSARs

Lipnick, R.L. (1993) Baseline toxicity QSAR models a means to assess mechanism of toxicity for aquatic organisms and mammals, in... 

Morall, S.W., M.J. Rosen, Y.P. Zhu, D.J. Versteeg, and S.D. Dyer. 1997. Physical chemical descriptors for the development of aquatic toxicity QSARs for surfactants, Proceedings of the 7th Internatl. Workshop on QSARs in Environmental Science, SETAC Press, USA, in press. 

In some instances, researchers are able to develop separate Hansch equations for both activity and toxicity of a drug. Differing parameters between the activity and toxicity QSAR equations allow researchers to optimize properties that boost desired activity and minimize toxicity. The following two QSAR equations give the activity (12.d) and toxicity (12.e) for a series of nitrogen mustard antitumor compounds (see Chapter 6). Based on these equations, is it likely that a nontoxic nitrogen mustard will be developed Explain. 

Also important to the validation process of QSARs is vertical validation. In this instance, quantitatively similar QSARs are developed with similar descriptors but using data for a different toxic endpoint. For example, the investigation of Karabunarliev et al. (1996b) modeled acute aquatic toxicity data for the fathead minnow Pimephales promelas. The compounds considered in the analysis were confined to substituted benzenes, and descriptors limited to log Kow and Amjx. The fish toxicity QSAR (log [LQ,]-1 = 0.62 log K, + 9.17 A - 3.21 n = 122 R2 = 0.83 i = 0.16 F = 292) of Karabunarliev et al. (1996b) was very similar in terms of slope, intercept, and statistical fit to the QSAR presented in Equation 12.2. The fact that different endpoints provide very similar QSARs indicates that the QSAR is valid across protocols. This shows the universality of the model. 

Receptor-Mediated Toxicity QSARs for Estrogen Receptor Binding and Priority Setting of Potential... 

To date a number of different QSAR models have been published, for the prediction of drag toxicity QSAR models cover over 30 different endpoints. It is not within the scope of this review to cover all of them here. Many authors discussed and evaluated different QSAR models in the literature (Dearden 2003 Patlewicz et al. 2003 Tuppurainen 1999). 

R. L. Lipnick, in QSAR in Toxicology and Xenobiochemistry, M. Tichy, Ed., Elsevier, Amsterdam, 1985, pp. 39—52. Validation and Extension of Fish Toxicity QSARs and Interspecies Comparisons for Certain Classes of Organic Chemicals. 

McCarty LS, Mackay D, Smith AD, Ozburn GW, Dixon DG. Interpreting aquatic toxicity QSARs The significance of toxicant body residues at the pharmacologic endpoint. Sci Total Environ 1991 109-10 515-25. 

Cronin, M.T.D. and Worth, A.P. (2008) (Q)SARs for predicting effects relating to reproductive toxicity. QSAR Comb. Scl, 27, 91-100. 

Robert D, Carbo-Dorca R. Aromatic compounds aquatic toxicity QSAR using quantum similarity measures. SAR QSAR Environ Res 1999 10 401-422. 

Figure 2. Baseline narcosis aquatic toxicity QSAR model. For nonelectrolytes acting by a narcosis mechanism, no toxicity is observed if the predicted toxic concentration exceeds the water solubility. With decreasing test duration, pseudo-steady-state partitioning is not achieved for very hydrophobic chemicals, and the location of the water solubility cutoff shifts to chemicals having a lower partition coefficient. Compounds acting by more specific mechanisms produce toxicity at lower aqueous concentrations than predicted by narcosis, and fall within the domain of excess toxicity.

Calculations of Tg adapted from refs. 63 and 88, which were based upon linear or bilinear baseline toxicity QSAR models. 

Tixier et al. [4] have identified, synthesized, and assessed the toxicity of all transformation products of diuron. The bio assay they used was a bioluminescence inhibition test with the marine bacterium Vibrio fischert Since diuron does not exhibit any specific mode of toxic action towards bacteria, the QSAR analysis using a rescaled QSAR for Vibrio fischeri [42] only confirmed that diuron and all its metabolites with the exception of DCA (3,4-dichloroaniline) act as baseline toxicants (Table 3, for full names of metabolites see Fig. 4). However, DCA was 46-times more toxic than predicted with the baseline toxicity QSAR and almost two orders of magnitude more toxic than the parent compound. Such a specific mode of toxic action of a transformation product cannot easily be predicted unless toxicophores like the aniline structure present in DCA are considered as a signal. This is discussed in the conclusion section in more detail. 

Ownby, D.R., and M.C. Newman. 2003. Advances in qnantitative ion character-activity relationships (QICARS) Using metal-ligand binding characteristics to predict metal toxicity. QSAR Comb. Sci. 22 241-246. 

Lepadatu, C., M. Enache, and J.D. Walker. 2009. Toward a more realistic QSAR approach to predicting metal toxicity. QSAR Comb. Sci. 28 (5) 520-525. 

Acute toxicity, chronic toxicity, QSAR, bioconcentration, chemical potency, toxicokinetics, organic chemicals. 

To quantify and discuss the relationship between acute and chronic toxicity QSARs, bioconcentration, and the descriptor octanol/water partition coefficient (log Kq ) for some organic chemicals causing toxicity, primarily by narcosis or narcosis-like modes of action. 

A variety of toxicity QSARs obtained from literature sources has been recalculated using the geometric mean regression technique and is presented in TABLET... 

For recalculated QSARs (equations through 14 in TABLE 1), it appears that both acute and chronic toxicity QSARs can be described by a one-constant equation since the slope of the regression, in most cases, appears to be approximately unity specifically, negative one. Equations 12 and 14 do not appear to follow this relationship as some regression coefficients are substantially different from unity. This may not be a fundamental difference and may be attributable to the lack of proper ionization comperisation. Nonetheless, the regression intercepts for the phenolics appear to be nearly an order of magnitude lower than those for the equivalent narcotic QSAR. 

McCarty et al. (1985) discussed combining the known relationships between acute and chronic toxicity QSARs and Kqw as well as between Kg and Kq, . The general character of the relationship between toxicity, bioconcentration, and octanol/water partition coefficient is readily apparent upon inspection of FIGURE 1, which presents equations 5, 9 and 11 for narcotic organic chemicals and fathead minnows. The relationship between acute and chronic toxicity QSARs is also clearly illustrated. 

An ionization-corrected bioconcentration relationship for rainbow trout is not available. However, using the uncorrected toxicity QSARs with the narcotics bioconcentration relationship (equation 5) may produce reasonably valid constants as both relationships are based on total uncorrected water concentration. When the toxicity and bioconcentration relationships are combined, the errors due to ionization may cancel out. This procedure generates, from equations5and 12aand b, and 14aand b,constantsofO.042 and 0.31 mmol L (acute) and 0.033 and 0.22 mmobL" (chronic), respectively. 

The above analysis also answers the question of whether acute and chronic toxicity QSARs are parallel. Since the regression coefficients are all essentially unity for the cases presented (exceptions noted), they do appear to be parallel. 

Acute and chronic toxicity QSARs, for the toxicants and species reported herein and using log as the descriptor, appear to be parallel, each having a slope near unity. 

Equation 1 (TABLE 2) is the toxicity QSAR for this data set a slope of essentially negative one and r of 0.99 is achieved. Equation 2 is generated using equation 1 and the bioconcentration/Kow relationship of Halfon (1985). It indicates that a whole-body toxicant concentration of approximately 6,500 yumol L or 0.0065 mol L (or mol Kg when the density is about 1.0) is associated with an acutely toxic lethal response in half the exposed population of fathead minnows at essentially infinite time, i.e., threshold. 

---