Abstract
This PhD thesis takes place in a context of the contamination of marine environment by emerging contaminants, such as pharmaceutical active compounds (PhAC), and their effects on marine organisms. The study of causal relationships between exposure to one or more contaminants and the response of the Mediterranean mussel, Mytilus galloprovincialis, was the central focus of the research presented in this thesis. In order to provide information on the mechanisms of action and potential toxic effects of PhAC, data lacking in the literature, -omic approaches such as metabolomics and proteomics were applied. The effects of the antiepileptic carbamazepine (CBZ), a PhAC frequently detected in the marine environment, were first investigated on M. galloprovincialis through an integrated approach of metabolomics and proteogenomics. The data fusion strategy applied revealed correlated protein and metabolic signatures in response to exposure. The use of bioinformatics tools that put proteins and metabolites into their biological context thus highlighted changes in protein synthesis, fatty acid degradation, amino acid and carbohydrate metabolism, and cell death programming. Although the study of the effects of a single contaminant is essential to obtain mechanistic information, it is far removed from the relevance of environmental exposure, since organisms are exposed simultaneously to a multitude of contaminants. In order to integrate this complexity, mussels M. galloprovincialis were exposed to a wastewater treatment plant (WWTP) effluent, the main pathway of PhAC into the marine environment. Analysis of the metabolic fingerprints generated was first performed on male mussels to rule out gender-related biological variability (which could hide the response to exposure). Several metabolic pathways were thus revealed to be impacted (e.g. amino acids, purines and pyrimidines, Krebs cycle, neurohormones, etc.) which can disrupt several biological functions and processes (e.g. energy metabolism, immune system, osmorregulation, reproduction, byssal formation, etc.) and have adverse consequences on the organism. Based on the literature, hypotheses of causal relationships have been established between certain contaminants detected in the WWTP effluent (38 PPs and 4 pesticides) and the effects observed. Based on the same experiment, the gender factor was then integrated into the processing of data acquired from male/female and exposed/unexposed individuals in order to understand the role of gender in the response to exposure. To this end, the statistical approach of Analysis of Variance Multiblock Orthogonal Partial Least Square proved to be relevant for this kind of multifactorial experimental design. This approach was thus able to characterize and relate the variability of metabolomics data to its different factors of origin. A common response between the two genders, related to the exposure factor, was demonstrated through the modulation of several lysophospholipids induced by oxidative stress. While a gender-specific response, related to the interaction between gender and exposure factors, has been described following a modulation of certain polar lipids according to gender and a disruption of the kynurenin pathway only in males. This thesis work was able to strengthen knowledge on the effects of a PhAC of concern for the environment, CBZ, excluded from any regulatory framework, as well as on the effects of an exposure close to the environmental conditions reconstituted through a WWTP effluent. Original approaches to effects investigation and data analysis have been pertinently applied.