Computational Analysis and Modeling of Tox21 AhR Activation High-Throughput Screening Assays Público
Zhao, Youni (Spring 2021)
Abstract
Dioxin-like compounds (DLCs) including the most potent 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) are important environmental pollutants that can cause potential harms to human health including endocrine disruption and immunosuppression. The toxicities of DLCs are mainly mediated by the aryl hydrocarbon receptor (AhR) in the cytoplasm, which after ligand binding enters the nucleus to bind to dioxin response elements (DRE) to alter the transcription of a number of genes including xenobiotic-metabolizing enzymes and many others. Induction of AhR-mediated genes has been demonstrated to be highly nonlinear in many cases, including steeply sigmoidal responses and nonmonotonic dose responses (NMDR). The present study aimed to analyze the Tox21 AhR high-throughput screening assay data to study the dose-response relationships of more than 10K chemicals. After applying a customized unsupervised and supervised machine learning approach and screening against the parallel cell viability and luciferase interference assays, 1380 active chemicals were selected and their concentration-response curves were classified into 6 shape categories: monotonic increasing, monotonic decreasing, flat, U, Bell and S shape. Among them, 2 are U shaped, 68 Bell shaped, 1092 monotonic increasing, 27 monotonic decreasing, and 1 S shaped. To investigate the nonlinearity of monotonic increasing curves, Hill function was fitted to the data and it was found that the majority of the curves have a Hill coefficient between 1-5, suggesting sigmoidal responses. To understand the potential mechanism of these sigmoidal responses, a stochastic mathematical model of positive cooperative binding between liganded AhR and DREs was constructed. The model demonstrated that with allosteric increase in binding affinity between 3, 5 and 7 DREs, the gene transcription can exhibit sigmoidal responses with Hill coefficient as high as 2.5. In summary, the present study demonstrated that AhR-mediated gene transcription induced by many chemicals can be highly nonlinear, which has significant implications in the risk assessment of these compounds.
Table of Contents
INTRODUCTION
1. Toxicity of TCDD and dioxin-like compound (DLCs)
2. Molecular initiating event (MIE) and mechanism of DLCs
3. Next-generation risk assessment (NGRA) using new approach methodology (NAM)
4. Tox21 program and EPA ToxCast program
5. Nonlinear and non-monotonic dose response (NMDR)
5.1 Dose-response relationship
5.2 Nonlinearity (ultrasensitivity) and NMDR of AhR-mediated responses
5.3 Tox21 assay as a valuable resource to study AhR-mediated nonlinear dose responses
METHODS
1. Tox21 AhR assay
2. Unsupervised machine learning to cluster concentration-response curves
3. Supervised machine learning to classify concentration-response into different shapes
4. Correlation analysis to rule out cytotoxicity and luciferase interference effects
5. Curve-fitting for Hill function
6. Computational modeling of AhR-DRE activation
RESULTS
1. Statistics of Tox21 AhR response assay and viability assay data and threshold values for chemical filtering
2. Unsupervised learning for concentration-response
3. Supervised results of each screening
4. Representative concentration-response curves in each category
5. Hill function fitting
6. Mathematical modeling of positive cooperative binding between AhR and DRE.
DISCUSSION
REFERENCES
About this Master's Thesis
School | |
---|---|
Department | |
Degree | |
Submission | |
Language |
|
Research Field | |
Palavra-chave | |
Committee Chair / Thesis Advisor |
Primary PDF
Thumbnail | Title | Date Uploaded | Actions |
---|---|---|---|
Computational Analysis and Modeling of Tox21 AhR Activation High-Throughput Screening Assays () | 2021-05-04 16:26:39 -0400 |
|
Supplemental Files
Thumbnail | Title | Date Uploaded | Actions |
---|