Transcriptomic signature comparisons identify conserved key events and respiratory disease signatures most similar to...

Wildfire events pose increasing threats to public health, especially concerning the respiratory system. The chemistries of emission mixtures vary widely based on factors including fuel source, burn temperature, and atmospheric and photochemical processes. Because of these varying exposure conditions, differing biological responses may occur, making it difficult to identify common mediators of wildfire smoke-induced disease with which to base health and risk assessments upon. To address this difficulty, this study set out to identify genes and modes of action highly shared across biological responses to wildfire-relevant exposures and respiratory disease outcomes. Specifically, we mined the Molecular Signatures Database and Computational Toxicogenomics Database for human consensus gene expression signatures of respiratory diseases and infections putatively linked to wildfire smoke exposures. These gene signatures were then compared to transcriptomic signatures measured in the lung of mice exposed to condensates derived from variable controlled biomass burn conditions. Data cleaning followed by similarity scoring, clustering, data reduction, and pathway enrichment methods were used to identify patterns amongst exposure and disease profiles. Fifty-three genes were identified as highly shared across signatures of both exposure and disease. Jaccard-based similarity scoring analyses of these top shared genes revealed asthma as the respiratory disease most similar across the majority of biomass conditions under evaluation. Pathway enrichment of these top shared genes identified interleukin signaling, oxidative stress response, and estrogen signaling as mechanisms of action likely linking exposure-to-disease outcomes across these variable smoke exposure conditions. Altogether, these results highlight consistent mediators to use in future health risk-based evaluations of wildfire smoke emissions.