The Deepwater horizon oil spill released over 210 million gallons of crude oil into the Gulf of Mexico. To protect the environment a multi-month effort was launched including the use of Corexit oil dispersants that brought adverse effects to those involved in the cleanup activities. The impact of the of oil-dispersant mixtures on the recovery prompted an interest to not only determine the biological effects of this mixture on humans, but to expand the efforts to identify potential biomarkers of exposure for future monitoring efforts. Recognizing the potential for oildispersant by-products to be inhaled, the objective of this study was to compare the expression profiles derived from an ex-vivo human lung exposure model BEAS2B exposed to the Water Accommodated Fraction (WAF) of oil vs. dispersant vs. oil-dispersant mixtures in a cell model, as a strategy to identify potential biomarkers of exposure. Ex vivo model was employed to evaluate BEAS2B whole genome expression patterns following exposure to WAF-dispersed oil samples, WAF-Crude, and Corexit dispersants 9500 and 9527. Differential expression analysis using DESeq2 and DAVID to annotate Gene Ontology, KEGG and SP-PIR allowed identifying up- and down-regulated genes and functional terms. Thirty-nine response genes under different treatments overlapped. From these results, we selected SERPINE2 as a potential gene candidate for further evaluation. SERPINE2 was selected as multiple studies demonstrate an association with respiratory disorders. A meta-analysis was performed to systematically assess the association between the rs3795879 G/A polymorphism in SERPINE2 and the development of respiration disorders. Rs3795879 is in complete linkage disequilibrium with other genetic variant nearby, it has an r2 value of 1 in HapMap. The odds of frequency G vs A is 0.901 in cases and controls with 95%CI (0.810,1.003), suggesting that the relationship is not significant. Further laboratory and epidemiological studies are needed to validate the potential use of this marker. Our study demonstrates that whole genome expression studies constitute a valuable resource for the potential identification of biomarkers.
