Quantitative structure—activity relationships for the prediction of relative in vitro potencies (REPs) for chloronaphthalenes

Chloronaphthalenes (CNs), due to their structural similarities to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and the other “dioxin-like” compounds, can bind to the aryl hydrocarbon receptor (AhR) and induce a wide range of pleotrophic effects. Relative potency of individual dioxin analogues can be measured relative to that of TCDD. Relative effects potencies (REP) can be based on many responses, including in vivo and in vitro responses. Both in vivo and in vitro tests, based on either indigenous responses such as the induction of ethoxyresorufin O-deethylase (EROD) or exogenous reporter genes under the control of the AhR such as luciferase can be used to determine REP values. Here we used measured REP values determined for CNs in two assays. Both assays are based on H4IIE rat hepatoma cells. The H4IIE assay is based on expression of the endogenous reporter gene (CYP-1 A) that codes for the expression of EROD and the H4IIE-luc assay which is based on the exogenous reporter gene (luciferase) transfected into the H4IIE cell line. Experimentally determined REP were available for only 17 and 18 of the 75 possible choronaphthalene congeners, for the H4IIE and H4IIE-luc assays, respectively. For this reason computational models were developed to allow prediction of the relative potencies of the other CN congeners. Predictive relationships were based on quantum chemical descriptors obtained from Density Functional Theory (DFT) calculations (B3LYP/6–311++G∗∗). The final models were found by means of a hybrid method combining a genetic algorithm and artificial neural networks. REP values estimated for individual CNs based on the H4IIE assay ranged from 4.3 × 10^{− 9} to 3.2 × 10^{− 2} while those based on the H4IIE-luc assay ranged from 4.0 × 10^{− 8} to 1.8 × 10^{− 3}. CN congeners nos. 66, 67, 70 and 73 were exhibited the greatest REP values in both assays. The 1,2,3,5,6,8-hexaCN congener (no. 68) had a REP value that was 10-fold less. The remaining congeners had REP values that were less or did not cause sufficient up-regulation of the monitored genes to allow for the calculation of a REP. Interactions of CNs with the AhR could be affected by three possible factors: molecular size, steric interactions and electrostatic interactions. These findings are discussed relative to the use of consensus TCDD equivalency factors' (TEFs) for use in risk assessments of CNs for regulatory purposes.

Keywords:

• Chloronaphthalenes
• REPs
• EROD
• H4IIe
• H4IIE-luc
• luciferase
• QSAR
• dioxin-like compounds
• PCNs

Acknowledgments

This study was supported by the Ministry of Education and Science under Grant no. KBN 1128/T09/2003/24 and DS/8250-4-0092-6. Computations were conducted using computers in the Academic Computer Center in Gdańsk TASK. Dr. Tomasz Puzyn is the recipient of a fellowship from the Foundation for the Polish Science.

Notes

∗High uncertainty due to extrapolation outside of the model's domain.

^aVilleneuve et al.^{[ 17 ]}

^bVilleneuve et al.^{[ 49 ]}

^cFalandysz and Puzyn^{[ 30 ]}

^dThis study.

^eBlankenship et al.^{[ 15 ]}

^fOlivero-Verbel et al.^{[ 31 ]}; I = Inactive, A = Active.

^aTopological descriptor;

^bquantum-chemical descriptor calculated in vaccuo (B3LYP/6–311++G∗∗);

^cquantum-chemical descriptor calculated in water (PCM model, B3LYP/6–311++G∗∗);

^dquantum-chemical descriptor calculated in octanol (PCM model, B3LYP/6–311++G; dielectric constant ε_oct = 10.3, solvent radius r = 3.250 Å); ∗A data matrix presenting values of 40 molecular descriptors calculated for the 75 possible CN congeners is available from the corresponding author.

∗Indicates greater than 1 order of magnitude of uncertainty between assay systems. The greater of the two values was used for the TEF.