Integration of multi-omics and benchmark dose modeling to support adverse outcome pathways.

Background: Recent advancements in omics and benchmark dose (BMD) modeling have facilitated identifying the dose required for a predetermined change in a response (e.g. gene or protein change) that can be used to establish acceptable dose levels for hazardous exposures. Adverse Outcome Pathways (AOPs) describe the causal links between toxicants and adverse effects through key events (KEs). Integrating omics data within the AOP framework quantitatively links early molecular events to later phenotypic effects. In this study, we use omic-based BMD analyses in an in vitro blood model exposed to radiation to identify point of departure (POD) values across KEs to acute myeloid leukemia (www.aopwiki.org/aop/432).

Methods: Isolated white blood cells were cultured and X-irradiated (1 Gy/minute, 0-6 Gy). Transcriptomic and proteomic changes were assessed 24 h post-exposure. BMD modeling was applied and significantly perturbed genes/proteins and pathways were identified. Those pathways relevant to KEs outlined in AOP 432 were grouped and a POD was determined.

Results: BMD modeling identified 1294 genes and 167 proteins with median BMD lower confident limit (BMD) values of 1.35 and 0.32 Gy, respectively. Pathway analysis identified biological processes related to DNA damage/repair, oxidative stress, cell cycle regulation, immune responses, and cancer development. These findings aligned with the KEs in AOP 432. The BMDL values of canonical pathways associated with these KEs were generally below 0.5 Gy with specific genes (e.g. GADD45A) displaying BMDLs <0.05 Gy.

Conclusions: This work provides insights into predictive radiation induced mechanisms and associated dose of activity that can be taken into consideration for low dose (< 0.1 Gy) risk analysis.