PhD thesis defense - Corentin MIT

Date de publication : Nov 15, 2019 9:55:40 AM

Abstract:

Biomarkers are useful tools for the diagnosis of environmental risk in aquatic ecosystems. Nevertheless, the measurement of these sub-individual markers still presents some limitations for the assessment of ecosystem health, including the characterisation of the complex dynamics of responses of these non-lethal effects as a function of time or dose, or the extrapolation of responses from one scale of biological organisation to another. One of the solutions that seems promising for characterising the dynamics of these responses from a change of scale perspective would be to integrate the biomarkers into mechanistic models that make it possible to predict these dynamics and explain the mechanisms underlying the effects. This thesis proposes to build mechanistic models of physiologically based toxicokinetics and toxicodynamics (PBTK-TD) to better characterise and understand the response dynamics of biomarkers. In this context, the problem of biomarker dynamics was divided in two. First, the "toxico-kinetic" or TK makes it possible to link the external dose, present in the environment, to the internal dose, present in the organism. Second, the "toxico-dynamic" or TD, makes the link between the internal dose and the effect. Accordingly, the first step in this thesis was to collect a set of TK and TD data in our model species, the three-spined stickleback, on a family of compounds, the bisphenols, and more specifically, BPA, BPS and BPF. These data, collected from short-term exposures (seven days of contamination and seven days of depuration) and long-term exposure (21 days), were used to compare the modulating effects of bisphenols on biomarkers. In particular, markers of innate immunity were strongly impacted by these substances. Differences in kinetics between BPA and BPS were also highlighted. Subsequently, the data collected during the exposures were used to build a physiologically based TK model (PBTK) for BPA, then a PBTK model coupled with TD sub-models (PBTK-TD) describing the dynamics of certain immunomarkers in the stickleback. Finally, a last PBTK-TD model was built to demonstrate the feasibility of this modelling approach for integrating exposure conditions more representative of those in the natural environment, i.e. for a mixture of substances. Taken as a whole, this thesis demonstrates the attractiveness of coupling the experimental approach consisting in measuring biomarkers and modelling.