Résumé
The negative impact on health of atmospheric particles (aerosols or particulate matter, PM) at the pulmonary, cardiovascular, and neurological levels is widely recognized. However, the understanding of the PM toxicity emitted from different emission sources and the links with their chemical composition are still poorly known. This is especially the case for sources mainly involved in PM pollution episodes (ie. > 50 µg m-3 of PM10 for several consecutive days). In France, and also in Europe, secondary formation of ammonium nitrate and sulfate (late winter -early spring), biomass burning (in winter, mainly due to residential wood heating), and desert dust (e.g. Saharan dust impacting the French West Indies), account for the main contributors during such episodes (Favez et al., 2021). In such context, the main objective of the SOTOX project is to make a comparison of the toxic potential induced by the predominant sources involved in the episodes of particulate pollution observed in France. Biomass combustion sources (primary emissions) and Saharan dust are therefore targeted in the first place. The originality of the work is to propose a differentiated assessment of the toxic potential of various PM sources based on acellular, biological (cellular in vitro) tests, together with an extensive chemical characterization, using PM samples collected during experiments performed in simulated real-world conditions or in ambient air.A set of PM samples collected on filters (n = 80), through different research projects allowed covering the following PM emissions sources:(a) Primary emissions from residential heating: open fireplace, old (5*) logwood stove, modern (7*) logwood/pellet stoves/boilers, oil boiler; tested under different output conditions (nominal and reduced) and using different wood species (hard and softwood),(b) Garden green waste burning (falling leaves, hedge trimming),(c) Desert dust (Saharan episodes in Martinique).The intrinsic PM toxicological potential (/ µg of PM) has been assessed by measuring different toxicity indicators from filter extracts (organic solvent or artificial pulmonary fluid). The acellular oxidative potential (OP) of the PM was determined using different specific probes (ascorbic acid (AA) and dithiothreitol (DTT)). Oxidative/nitrosative stress induction was investigated on human lung cells (A549) using the H2DCFDA (2′,7′-Dichlorodihydrofluorescein diacetate) probe. A bio-analytical approach based on the measurement of the activation of the AhR receptor (aryl hydrocarbon receptor) in human liver cells (HepG2) exposed to PM extracts permitted to evaluate the content of the particles in dioxin- and HAP-like pollutants. In addition, the analysis of key particulate toxic species (metals, polycyclic aromatic hydrocarbons (PAHs)) and their nitrated and oxygenated derivatives), as well as the PM carbonaceous content (elemental and organic carbon, EC-OC), has been performed to study the links between toxicity indicators and PM chemical composition. Results obtained will be compared in terms of intrinsic PM toxicological potential according to the sources, toxicity indicators and PM chemical composition.This work has been performed within the framework of the SOTOX project funded by the French air quality penalty