Prediction of zebrafish embryonic developmental toxicity by integrating omics with adverse outcome pathway

Xiao Gou1, Cong Ma1, Huimin Ji1

  • 1State Key Laboratory of Pollution Control & Resource Reuse, School of the Environment, Nanjing University, Nanjing 210023, Jiangsu, China.

Insights

A new computational toxicology method, ScoreAOP, effectively predicts chemical developmental toxicity in zebrafish embryos using adverse outcome pathways (AOPs) and omics data. ScoreAOP accurately identifies defects and elucidates mechanisms, outperforming methods focused solely on molecular initiation events (MIEs).

Area of Science:

  • Computational toxicology
  • Omics-based high-throughput bioassays
  • Adverse Outcome Pathways (AOPs)

Background:

  • New Approach Methodologies (NAMs) provide mechanistic insights (MIEs, KEs) but applying this to predict Adverse Outcomes (AOs) remains challenging.
  • Integrating omics data with AOP knowledge is crucial for advancing predictive toxicology.
  • Current methods struggle to bridge the gap between molecular events and organism-level toxicity predictions.

Purpose of the Study:

  • To develop and evaluate ScoreAOP, an integrated method for predicting chemical developmental toxicity in zebrafish embryos.
  • To assess ScoreAOP's ability to utilize AOPs and dose-dependent transcriptome data for toxicity prediction.
  • To compare ScoreAOP's performance against ScoreMIE, a method focused on MIEs.

Main Methods:

  • ScoreAOP integrates four AOPs and dose-dependent reduced zebrafish transcriptome (RZT) data.
  • Key rules for ScoreAOP include sensitivity of key events (PODKE), evidence reliability, and KE-AO distance.
  • Eleven chemicals with diverse modes of action were tested to evaluate ScoreAOP's predictive power and mechanistic insights.

Main Results:

  • ScoreAOP accurately predicted developmental defects for all eleven tested chemicals.
  • Eight of eleven chemicals showed developmental toxicity in apical tests.
  • ScoreAOP outperformed ScoreMIE, correctly predicting 11 vs. 8 chemicals, and successfully clustered chemicals by mode of action.

Conclusions:

  • ScoreAOP is a promising approach for applying omics-derived mechanistic information to predict chemical-induced Adverse Outcomes (AOs).
  • The method provides mechanistic explanations, such as the role of aryl hydrocarbon receptor (AhR) activation in cardiovascular dysfunction leading to developmental defects.
  • ScoreAOP advances computational toxicology by bridging molecular events to organism-level toxicity predictions.

Related Concept Videos