ABSTRACT: Human toxicogenomic studies to date have been of limited size, have mainly addressed exposures at the upper end of typical ranges of human exposure, and have often lacked precise, individual estimates of exposure. Previously, we identified genes associated with exposure to high (>10 ppm) levels of the leukemogen, benzene, through transcriptomic analyses of blood cells from small numbers of occupationally exposed workers. Here, we have expanded the study to 125 workers exposed to a wide range of benzene levels, including <1 ppm. Study design, and analysis with a mixed effects model, removed sources of biological and experimental variability and revealed highly significant widespread perturbation of gene expression at all exposure levels. Benzene is an established cause of acute myeloid leukemia (AML), and may cause one or more lymphoid malignancies in humans. Interestingly, acute myeloid leukemia was among the most significant pathways impacted by benzene exposure in the present study. Further, at most exposure levels, immune response pathways including T cell receptor signaling, B cell receptor signaling, and Toll like receptor signaling were impacted, providing support for the biological plausibility of an association between lymphoma and benzene exposure. We also identified a 16-gene expression signature modified by all levels of benzene exposure, comprising genes with roles in immune response, inflammatory response, cell adhesion, cell-matrix adhesion, and blood coagulation. Overall, these findings support, and expand upon, our current understanding of the mechanisms by which benzene may induce hematotoxicity, leukemia and lymphoma. Furthermore, this study shows that with good study design and analysis, transcriptome profiling of the blood of chemically-exposed humans can identify relevant biomarkers across a range of exposures and inform about potential associations with disease risks. Key words: occupational exposure 29 workers exposed to <<1ppm benzene at most dosimetry reading over a 14-month period (Very Low), 30 exposed to average <1ppm (Low), 11 exposed to 5-10 ppm (High), 13 exposed to >10 ppm (Very High), and 42 unexposed controls (Control) who were frequency-matched to these subjects on the basis of age and gender were analyzed (technical replicates on some subjects, see SubjectID/description).
