28 Assessment of Gene Expression Modulation Using Graph Analysis on Pairwise Expression Ratios. Impact of PM2.5-0.3 on 3-D Bronchial Epithelium Model.

Abstract Chronic exposure to fine particles (PM2.5-0.3) can have dramatic consequences for health, especially for the most vulnerable people, such as asthmatics. In order to better understand the impact of PM2.5-0.3 from different emission sources on the modulation of gene expression, a 3-D in vitro model of human bronchial epithelium (MucilAir-HF™) reconstructed from primary cells from healthy (EpiH) or asthmatic (EpiA) donors was used. Repeated exposures to PM2.5-0.3 (collected at industrial or road traffic sites) were performed during three weeks, and epithelia were sacrificed to extract RNAs and assess gene expression. Data were analysed according to emission sources, physiological status and exposure doses using an innovative method consisting of a graph analysis on pairwise expression ratios. Results obtained were related to those obtained with the common ΔΔCt method. Graph analysis on pairwise expression ratios proved to have better statistical properties than the common ΔΔCt method and demonstrated that repeated PM2.5-0.3 exposures induced a dose-dependent up-regulation of metabolic genes (CYPs) and a down-regulation of inflammation gene (CXCL10). These modulations were greater for "industrial" than for "urban traffic" PM2.5-0.3, and the effects were found to be greater after exposure of EpiA compared to EpiH, emphasing the importance of the epithelium physiological status in the sensitivity to particles. In conclusion, our study, original in terms of experimental conditions and statistical analysis method, highlights the importance of particle chemistry on the modulation of cellular and molecular responses, which may vary according to the vulnerability of the individuals.