The Development of an Adverse Outcome Pathway (AOP) Focusing on Immunosuppression by Glucocorticoid Receptor Activation for the Use in Chemical Risk Assessment

Abstract

Background: Chemical risk assessment and regulations by authorities are crucial to minimise adverse outcomes of hazardous chemicals on human health and the environment. Currently, toxicity testing of chemicals is facing a shift, moving away from traditional reliance on animal models to alternative (in vitro and in silico) approaches. Adverse outcome pathways (AOPs) are an established tool for hazard assessment and can structure and integrate data from a range of methodologies. By categorising mechanistic data into measurable biological events, causal links and areas for in vitro test development can be identified. AOP 14 (“Glucocorticoid Receptor Activation Leading to Increased Disease Susceptibility”) is an AOP covering immunosuppression and consists of seven biological events, five of which are not developed. Aim: This master thesis aimed to further develop AOP 14 by exploring evidence collection methods and examining the effects of glucocorticoids (GCs) on T cells. Methods: Two evidence collection methods were tested and compared to identify literature relevant to the development of AOP 14. A structured literature search focusing on reviews was conducted and compared to an artificial intelligence-based text mining approach, using the AOP-helpFinder. Results: The AOP-helpFinder retrieved 27 606 unique articles, approximately seven times more than the 3907 records from the literature search, both covering the entire AOP 14. The degree of overlap between the two methods varied (10% - 86%), depending on the precise comparison. Among the reviews retrieved by the literature search, 14 reviews described GC effects on T cells and associated cytokines, of which 12 were also identified by the AOP-helpFinder. Closer examination of the evidence indicated that physiological GC doses or long-term treatment, can cause a shift from Th1 to Th2 immune response. This was caused by a suppression of Th1 cytokines and interleukin-12, and a stimulation of Th2 cytokines. At pharmacological GC doses or acute treatment, both Th1 and Th2 cytokines were suppressed. Conclusion: This thesis significantly contributes to the development of AOP 14 by exploring two methods for AOP evidence collection. The integration of AI tools in AOP development, like the AOP-hF, is promising but it is still in its early stages and there is a need for thorough testing and evaluation prior to widespread implementation. Therefore, employing both approaches as complementary methods can enhance confidence in automated literature collection.