The use of molecular and genetic predictors and machine learning in assessing the toxicity of chemical substances

Abstract

Toxic hepatites is an important public health issue in Russia. They have diverse etiologies, including industrial toxins, drugs, and ethanol, leading to significant differences in pathogenesis at the molecular and genetic level. This study aimed to analyze genes involved in the response to toxic liver injury of different origins and to establish a model to predict the etiology of toxic hepatitis. To model acute toxic hepatitis, outbred white male rats weighing 180–200 g were used. They were divided into four groups: control group, tetrachloromethane group, paracetamol group, and ethanol group. 24 h and 72 h after toxicant administration rats were anesthetized and mRNA levels of Chek1, Gclc, Gstm1, Gstp1, Gstt1, Nfe2l2, Nqo1, and Ripk1 genes in liver homogenate were studied. Analysis of the expression of these genes showed that the expression profile varies depending on the etiology of toxic hepatitis. In tetrachloromethane poisoning, there was a marked increase in gene expression of Nqo1 (p?=?0.001), Gstm1 (p?=?0.037), and a decrease in gene expression of Nfe2l2 (p?=?0.004), Ripk1 (p?=?0.004). However, liver exposure to paracetamol and its metabolites resulted in decreased Gstm1 gene expression (p?=?0.001) and increased Nfe2l2 (p?=?0.009), Gclc (p?=?0.001), and Chek1 (p?=?0.011) gene expression. No significant changes in the expression of these genes are found in alcohol intoxication. A model was constructed to predict the etiology of toxic hepatitis on the basis of gene expression indices. Keywords: acute toxic hepatitis, paracetamol, tetrachloromethane, ethanol, gene expression.