Consideration of Styrene Transcriptomic Data Informs Mouse Lung Cyp2F2-Mediated Adverse Outcome Pathway

We read with considerable interest the recent paper in Hill and Conolly (2019). The authors describe a systematic review of research papers to organize an adverse outcome pathway (AOP) for mouse lung tumors following exposure to several compounds, including styrene, ethylbenzene, and fluensulfone, that all require metabolism by a mouse-specific cytochrome, Cyp2F2, to produce biologically active metabolites. The authors conclude that the unique prevalence of Cyp2F2-driven mouse lung tumors is due to differences in Club cell-dependent capacity for generating reactive metabolites compared with human lung. In the proposed AOP, Cyp2F2 generates reactive metabolites leading to cytotoxicity, cell death, proliferation, hyperplasia enhanced by toxicant-induced loss of Club cell-specific protein (CCSP) and eventually cancer. We and several coinvestigators have proposed a styrene mode of action (MoA) that was substantially informed by comparative whole genome transcriptomic responses in 2 wild-type mouse strains (CD-1 and C57Bl6), Cyp2F2-knockout mice, and transgenic mice lacking Cyp2F2 and expressing human Cyp2F1 (Andersen et al., 2017, 2018; Banton et al., 2019; Cruzan et al., 2018). These recently published MoA studies, which were not fully captured in the Hill and Conolly AOP, included exposures for 1, 5 days, and 4, 26, 52, 78, and 104 weeks. Like their proposed AOP, our mouse-specific MoA also starts with the key molecular initiating event of styrene metabolism to biologically active metabolite(s) by Cyp2F2. The essential role of Cyp2F2 metabolism and its lack of human relevance was evidenced by the complete attenuation of short- and long-term lung toxicity in Cyp2F2 knockout and Cyp2F1 humanized strains compared with wild-type mice. Importantly, however, the transcriptomic responses in wild-type mice at 1 day and 1 week across exposures ranging from 5 to 120 ppm showed enrichment in cell cycle, mitotic M-M/G1 phases, DNA synthesis and metabolism of lipids and lipoproteins. There were no signals for activation of stress pathways associated with oxidative injury or formation of reactive metabolites, activation of p53-mediated DNA-damage pathways, or those indicating altered glutathione metabolism. With longer-term exposures there were changes in a larger suite of transcription factors (TFs) associated with circadian control of cell metabolism that have been specifically linked to mouse terminal bronchiolar lung tumor development (Papagiannakopoulos et al., 2016), and also persistent downregulation of members of the NR4 family. Overall, the gene expression changes indicated that mouse-specific CYP2F2 metabolites affect nuclear receptor signaling controlling lipid metabolism, mitogenic cell proliferation, and associated loss of circadian metabolic control, and that reparative cell proliferation associated with cytotoxic reactive metabolites such as styrene oxide (inferred in the AOP) was not a key contributory MoA event. Our studies did not specifically examine whether the AOP-proposed loss of CCSP is a potential primary or associative key event in styrene-induced mitogenic cell proliferation. Other transcriptomic studies with compounds that cause cytotoxicity, eg, formaldehyde in the rat nose or chloroprene in mouse lung, show patterns of gene expression expected for these respective MoAs and the same patterns of response are seen with multiple exposures as with the shorter-term exposures (Andersen et al., 2010; Thomas et al., 2013). Unlike these and other reactive compounds examined, with styrene the prominent changes in pathways associated with cell cycle and lipid/cholesterol pathway response diminish at times beyond a week where the changes in TFs became more evident. Thus, the major difference between the styrene MoA informed by transcriptomic data and the proposed AOP is the hypothesized biological activity of the mouse-specific metabolites, ie, whether they are DNA-reactive and cytotoxic or biologically active signaling molecules activating TFs that mediate gene expression controlling proliferation and long-term circadian cycle disruption. Importantly, the transcriptomic data do not support involvement of DNA damage or repair pathways, challenging earlier MoA hypotheses that styrene oxide genotoxicity is a key driver of mouse lung carcinogenesis. We appreciate that the AOP proposed by Hill and Conolly (2019) is consistent with our MoA conclusion that these responses are mouse lung-specific, lacking qualitative and/or possibly quantitative relevance to humans since Cyp2F1 and/or other human lung Cyp enzymes have minimal to no metabolic activity compared with Cyp2F2 metabolism (as is the case for styrene). We applaud the effort to develop this AOP for compounds activated by Cyp2F2 and look forward to elaboration of the Hill and Conolly AOP with consideration of the extensive gene expression work conducted with styrene. The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Preparation of this letter was funded by separate Styrene Information and Research Center (SIRC) contracts to Andersen ToxConsulting LLC and Exponent, Inc. Although SIRC was provided an opportunity to review and comment on drafts of this letter, the final text and conclusions in the letter remain soleley those of the authors.