Mechanisms of olfactory toxicity of the herbicide 2,6-dichlorobenzonitrile: Essential roles of CYP2A5 and target-tissue metabolic activation

Abstract

The herbicide 2,6-dichlorobenzonitril (DCBN) is a potent and tissue-specific toxicant to the olfactory mucosa (OM). The toxicity of DCBN is mediated by cytochrome P450 (P450)-catalyzed bioactivation; however, it is not known whether target-tissue metabolic activation is essential for toxicity. CYP2A5, expressed abundantly in both liver and OM, was previously found to be one of the P450 enzymes active in DCBN bioactivation in vitro. The aims of this study were to determine the role of CYP2A5 in DCBN toxicity in vivo, by comparing the extents of DCBN toxicity between Cyp2a5-null and wild-type (WT) mice, and to determine whether hepatic microsomal P450 enzymes (including CYP2A5) are essential for the DCBN toxicity, by comparing the extents of DCBN toxicity between liver-Cpr-null (LCN) mice, which have little P450 activity in hepatocytes, and WT mice. We show that the loss of CYP2A5 expression did not alter systemic clearance of DCBN (at 25 mg/kg); but it did inhibit DCBN-induced non-protein thiol depletion and cytotoxicity in the OM. Thus, CYP2A5 plays an essential role in mediating DCBN toxicity in the OM. In contrast to the results seen in the Cyp2a5-null mice, the rates of systemic DCBN clearance were substantially reduced, while the extents of DCBN-induced nasal toxicity were increased, rather than decreased, in the LCN mice, compared to WT mice. Therefore, hepatic P450 enzymes, although essential for DCBN clearance, are not necessary for DCBN-induced OM toxicity. Our findings form the basis for a mechanism-based approach to assessing the potential risks of DCBN nasal toxicity in humans.

Introduction

The herbicide 2,6-dichlorobenzonitrile (DCBN, see Fig. 1 for structure), which is widely used for weed control (Cox, 1997), is one of the most potent olfactory toxicants in rodents (Brandt et al., 1990). In mice, DCBN treatment results in necrosis of the Bowman's glands, followed by degeneration and necrosis of the olfactory neuroepithelium. DCBN induces permanent loss of olfactory neurons in the dorsal medial part of the nasal cavity, accompanied by respiratory metaplasia in the damaged region (Bergman et al., 2002). Human exposure to DCBN occurs occupationally, through inhalation and dermal absorption, and in the environment, through ingestion of contaminated food and ground water.

The involvement of P450-catalyzed metabolic activation in the toxicity of DCBN has long been proposed, based on the substantial inhibition of DCBN-induced in vivo formation of DCBN-protein adducts, as well as olfactory toxicity, by metyrapone, a P450 inhibitor (Brandt et al., 1990, Brittebo, 1995). The olfactory toxicity of DCBN is believed to derive from its reactive metabolites (2,3-oxo-DCBN and, to a lesser extent, 3,4-oxo-DCBN), produced through P450-catalyzed metabolic activation; these epoxides readily form conjugates with reduced glutathione (GSH) or attack sulfhydryl groups in cellular proteins, yielding protein adducts (Ding et al., 1996). P450 enzymes in both liver, the main metabolic organ for circulating xenobiotic compounds, and the nasal mucosa, the target organ for DCBN toxicity, can metabolize DCBN in vitro. The potential role of target tissue metabolic activation in DCBN toxicity has been suggested, mainly on the basis of the unique occurrence of abundant, tissue-selective P450 enzymes, CYP2A5 and CYP2G1, in the mouse OM; both enzymes could metabolize DCBN, with relatively low km values, into reactive intermediates that form protein adducts and GSH conjugates (Ding et al., 1996, Gu et al., 1998). However, till now, there has been no direct proof regarding the tissue source (liver versus OM) of the reactive intermediates that are responsible for the olfactory toxicity of DCBN, and it is not known whether CYP2A5 or CYP2G1 is more important for DCBN toxicity in vivo. The latter question is important for attempts to extrapolate rodent toxicity data to humans, for risk assessment. Notably, while human CYP2A6 and CYP2A13, orthologs of mouse CYP2A5, are expressed in human nasal mucosa, and both enzymes are active toward DCBN metabolic activation (Liu et al., 1996, Su et al., 2000), human CYP2G1 has become a pseudogene (Sheng et al., 2000).

The aims of this study were to identify the P450 enzyme responsible for DCBN metabolic activation in vivo, and to determine whether hepatic P450-generated DCBN metabolites play a significant role in DCBN toxicity in the OM. Two knockout mouse models were utilized: the Cyp2a5-null mouse (Zhou et al., 2010) and the liver-Cpr-null (LCN) mouse (Gu et al., 2003). CYP2A5 is expressed in many tissues, including OM, liver, kidney, and lung. In the Cyp2a5-null mouse, the Cyp2a5 gene was deleted in the germline, and CYP2A5 expression is no longer detected in the OM or elsewhere. This model has been used to demonstrate a critical role of CYP2A5 in the systemic clearance of nicotine and cotinine (Zhou et al., 2010). In the LCN mouse, the gene for the NADPH-cytochrome P450 reductase (CPR) was deleted in a tissue-specific fashion, in all hepatocytes, leading to inactivation of all microsomal P450 enzymes, including CYP2A5. The LCN mouse has been used to clarify the role of hepatic CPR/P450 in the systemic clearance of various xenobiotics (e.g., Gu et al., 2007, Weng et al., 2007, Zhang et al., 2007); they have also been used to identify in vivo contributions of hepatic P450s to xenobiotic toxicity in extrahepatic target organs, such as the renal toxicity induced by acetaminophen (Gu et al., 2005).

In this study, the Cyp2a5-null, LCN, and wild-type (WT) mice, all on a C57BL/6 (B6) genetic background, were treated with DCBN at doses established in previous studies to cause nasal toxicity in WT mice. The extent of DCBN-induced OM toxicity was assessed by histological analysis of the nasal cavity, and through measurements of tissue levels of non-protein thiol (NPSH). The impacts of the global loss of CYP2A5 expression, or the hepatocyte-specific loss of CPR expression, on rates of DCBN metabolic activation in vitro, and on the kinetics of DCBN clearance in vivo, were also examined. The results of our studies provide definitive evidence that CYP2A5 plays an essential role in mediating DCBN toxicity in the OM, and that hepatic P450 enzymes, although essential for DCBN clearance, are not necessary for DCBN-induced OM toxicity.