Oxidative stress, glutathione, and CYP2E1 in 1,4-dioxane liver cytotoxicity and genotoxicity: insights from animal models

Abstract

1,4-Dioxane (DX) is an emerging drinking water contaminant worldwide, which poses a threat to public health due to its demonstrated liver carcinogenicity and potential for human exposure. The lack of drinking water standards for DX is attributed to undetermined mechanisms of DX carcinogenicity. This mini-review provides a brief discussion of a series of mechanistic studies, wherein unique mouse models were exposed to DX in drinking water to elucidate redox changes associated with DX cytotoxicity and genotoxicity. The overall conclusions from these studies support a direct genotoxic effect by high dose DX and imply that oxidative stress involving CYP2E1 activation may play a causal role in DX liver genotoxicity and potentially carcinogenicity. The mechanistic data derived from these studies can serve as important references to refine the assessment of carcinogenic pathways that may be triggered at environmentally relevant low doses of DX in future animal and human studies.

Introduction

1,4-Dioxane (DX) is a synthetic chemical stabilizer of chlorinated solvents, such as 1,1,1-trichloroethane (TCA) and trichloroethylene (TCE) [1]. Improper disposal of these industrial solvents is the primary source of DX release into the environment [1–5]. Due to its high water solubility, low degradability and poor adsorption, DX is persistent in the environment, being transported predominantly via water. Consequently, DX contamination of water bodies, including surface water and ground water, has been reported across the US and worldwide [6,7]. To date, there is a lack of human studies assessing the toxicity associated with oral DX exposure; however, studies in laboratory animals demonstrate that chronic DX exposure in drinking water induces liver adenomas and adenocarcinomas [8,9].

Despite concerns regarding its carcinogenic potential and the widespread contamination in the US, DX in drinking water is not regulated by the federal government and state-level regulatory guidelines vary greatly. This is, in part, due to the undetermined mechanisms of action (MOAs) of DX carcinogenicity, which represents a significant knowledge gap in DX risk assessment [10]. The mutagenic MOA of DX has been investigated and remains inconclusive, due to a lack of mutagenic activity in vitro and evidence of liver genotoxicity in vivo [10,11]. Utilizing several unique mouse models, our group performed a series of DX exposure studies involving low to high DX doses and short-term to subchronic exposures [12,13]. The objective of this mini-review is to summarize the main findings from these mechanistic studies and discuss new lines of experimental evidence that support a direct genotoxic effect by high dose DX and a potential causal role of oxidative stress in this process.

Multi-omics study in BDF-1 female mice reveals altered oxidative stress defense systems accompanying DNA damage as early changes by oral exposure to high dose DX

Oxidative stress occurs when reactive oxygen species (ROS) are produced at levels exceeding those capable of being sequestered by cellular antioxidant defense systems, which comprise small molecule antioxidants (e.g. glutathione (GSH)), antioxidant enzymes (e.g. glutathione S-transferases (GSTs)) and their regulators (e.g. nuclear factor erythroid 2-related factor 2 (NRF2)) [14]. Many drugs, environmental toxicants and selected dietary components cause liver toxicity by inducing oxidative stress [15–17]. Hepatic oxidative damage can result in non-neoplastic lesions (cytotoxicity), such as fatty liver (steatosis), hepatocyte death and inflammation [18], as well as preneoplastic and neoplastic lesions (carcinogenicity) [19]. Oxidative stress is an important mutagenic mechanism for numerous chemical carcinogens [20,21]. Prior to our studies, the involvement of oxidative stress in DX liver toxicity had been investigated in a limited number of in vivo studies (Table 1). Mnaa et al. reported increased lipid peroxidation in the livers of rats exposed to DX at 100 mg/kg/day in drinking water for 6 weeks [22]. Gi et al. reported no difference in liver 8-hydroxy-2’ -deoxyguanosine (8-OHdG) levels (an index of oxidative DNA damage) measured by HPLC-ECD in male gpt delta transgenic F344 rats following DX exposures up to 5000 ppm in drinking water for 16 weeks [23]. However, the same research group later identified that 8-oxo-dG (a form of oxidative DNA modification) was one of the three DNA adducts characteristic of DX treatment in livers from male F344 rats by 200 ppm or 5000 ppm DX for 16 weeks [24]. In this later study, DNA adducts were analyzed using more advanced HPLC-QTOF-MS.

Table 1.

In Vivo 1,4-Dioxane Exposure Studies Investigating Oxidative Stress and Genotoxicity in the Liver

Animal Model	DX Exposure Regime	Main Findings	Reference
Female Sprague-Dawley rats	100 mg/kg in drinking water for 6 weeks	Liver cytotoxicity  • Histopathology (portal edema, Kupffer cells activation, small focal hepatic necrosis associated with inflammatory cells infiltration)  • Increased serum ALP Liver genotoxicity  • No data reported Oxidative stress  • Increased marker of lipid peroxidation (MDA)  • Reduced catalase activity	Mnaa et al. (2016) [22]
Male gpt delta transgenic F344 rats	0.2, 2, 20, 200, 1000 or 5000 ppm in drinking water for 16 weeks	Liver cytotoxicity  • No treatment-related histopathology Liver genotoxicity  • Increased A:T- to -T:A transversion frequency by 1000 ppm  • Increased gpt mutation frequency, A:T- to -G:C transition and A:T- to -T:A transversion frequencies by 5000 ppm  • Increased mRNA level of DNA repair enzyme MGMT by 5000 ppm Liver Proliferation/preneoplastic lesion  • Increased mRNA level of PCNA by 5000 ppm  • Increased GST-P-positive foci by 5000 ppm Oxidative Stress  • No change in mRNA levels of 12 CYP enzymes  • No difference in 8-OHdG levels (measured by HPLC-ECD)	Gi et al. (2018) [23]
Male WT F344 rats	2, 20, 200, 2000, or 5000 ppm in drinking water for 16 weeks	Liver cytotoxicity  • No treatment-related histopathology Liver Proliferation/preneoplastic lesion  • Increased GST-P-positive foci by 2000 and 5000 ppm  • Increased BrdU labeling index by 5000 ppm	Gi et al. (2018) [23]
Male WT F344 rats	20, 200, or 5000 ppm in drinking water for 16 weeks	(Liver samples from Gi et al. 2018 were used for analyses)Liver genotoxicity  • Identified three DNA adducts (measured by HPLC-QTOF-MS) characteristic of DX treatment by 200 and 5000 ppm Oxidative Stress  • 8-oxo-dG is one of the three treatment-specific DNA adducts by 200 and 5000 ppm	Totsuka et al. (2020) [24]
Female BDF-1 mice	50, 500, and 5000 ppm in drinking water for 1 or 4 weeks	Liver cytotoxicity  • No gross histopathological change Liver genotoxicity  • Increased γH2AX-positive hepatocytes (DNA damage marker) by 5000 ppm at 1 and 4 weeks Oxidative Stress  • DEGs (measured by RNASeq analysis) enriched in GSH-mediated detoxification and NRF2-mediated oxidative stress response pathways by 5000 ppm at 4 weeks  • DAMs (measured by metabolomics analysis) enriched in metabolic pathways of nucleotide salvage synthesis and oxidative stress response by 5000 ppm at 4 weeks	Charkoftaki et al. (2021) [12]
Male B6 GclmKO and WT mice	1000 mg/kg/day by oral gavage for 1 week or 5000 ppm in drinking water for 12 weeks	Liver cytotoxicity (no difference between GclmKO and WT mice)  • Mild histopathology (single cell death and associated inflammation, ductal reactive change, and steatosis) at 12 weeks  • 2-fold increase in plasma ALT and AST activities at 12 weeks Liver genotoxicity (enhanced in GclmKO relative to WT mice)  • Increased 8-OHdG levels (by ELISA assay) at 12 weeks  • Elevated γH2AX/H2AX ratio (DNA damage repair marker) at 12 weeks Oxidative Stress (enhanced in GclmKO relative to WT mice)  • Transient induction of some NRF2-target antioxidant genes at 1 week  • Reduced GSH/GSSG ratio (index of oxidative stress) at 12 weeks  • Progressive induction of CYP2E1 and lipid peroxidation	Chen et al. (2022) [13]
Male B6 Cyp2e1KO mice	5000 ppm in drinking water for 1 week or 12 weeks	Liver cytotoxicity  • No treatment-related histopathology at 12 weeks  • No change in plasma ALT and AST activities at 12 weeks Liver genotoxicity  • Trend of increase in 8-OHdG levels (by ELISA assay) at 12 weeks  • Decreased γH2AX/H2AX ratio (DNA damage repair marker) at 12 weeks Oxidative Stress  • No induction of NRF2-targeted antioxidant genes  • No change in GSH/GSSG ratio (index of oxidative stress)  • No induction of lipid peroxidation	Wang et al. (Manuscript in preparation)

Abbreviations: 8-OHdG, 8-hydroxy-2′-deoxyguanosine; ALP, alkaline phosphatase; ALT, alanine transaminase; AST, aspartate aminotransferase; BrdU, bromodeoxyuridine; CYP, cytochrome P450; CYP2E1, cytochrome P450 2E1; DAMs, differentially abundant metabolites; DEGs, differentially expressed genes; DX, 1,4-dioxane; ELISA, enzyme-linked immunosorbent assay; Gclm, glutamate-cysteine ligase modifier subunit; GSH, glutathione; GSSG, glutathione disulfide; GST-P, glutathione S-transferase placental-form; H2AX, H2A histone family member X; γH2AX, phosphorylation of the Ser-139 residue of H2AX; HPLC-ECD, high performance liquid chromatography with electrochemical detection; HPLC-QTOF-MS, high-performance liquid chromatography-quadrupole time-of-flight mass spectrometer; MDA, malondialdehyde; MGMT, O6-methylguanine-DNA methyltransferase; NRF2, nuclear factor erythroid 2-related factor 2; PCNA, proliferating cell nuclear antigen; WT, wild type.

We performed a short-term exposure study in female BDF-1 mice using a range of DX doses (50, 500 or 5000 ppm) in drinking water for one or four weeks (Table 1) [12]. This study aimed to capture early molecular changes by DX in the liver by transcriptomic and metabolomic profiling along with histopathological examination. We chose the BDF-1 mouse strain because it has shown consistent evidence of liver tumors in the Kano et al. chronic study [25]. Livers from DX-exposed BDF-1 mice showed no gross pathological changes at any exposure dose or duration. However, the percentage of hepatocytes stained positively for phosphorylation of the Ser-139 residue of the histone variant H2AX (γH2AX), a marker of DNA damage, were elevated following one or four weeks of exposure to high dose DX (5000 ppm). Liver RNASeq analysis identified 65 differentially expressed genes (DEGs) by high dose DX for four weeks; top canonical pathways enriched by DEGs included xenobiotic metabolism, GSH-mediated detoxification, and NRF2-mediated oxidative stress response. Liver metabolomics in these same mice identified 81 differentially abundant metabolites (DAMs). These DAMs were involved in metabolic pathways of nucleotide salvage synthesis, oxidative stress response and detoxification, DNA damage, lipid metabolism, bile acid biosynthesis, and farnesoid X receptor (FXR)/ retinoid X receptor (RXR) activation. Collectively, we demonstrated that, in the absence of liver cytotoxicity, short-term exposure (up to four weeks) to high dose DX led to a mild increase in liver DNA damage and a hepatocellular oxidative stress response.

Deficiency in antioxidant GSH sensitizes mice to oxidative DNA damage following subchronic oral exposure to high dose DX

GSH is the most abundant cellular non-protein thiol, attaining millimolar concentrations in the liver [26]. GSH serves to protect cells against toxicity arising from exposure to excessive amounts of endogenous and exogenous electrophiles [27]. GSH scavenges hydroxyl radicals and superoxide directly, and serves as a cofactor for the enzyme glutathione peroxidase (GPX) in metabolizing hydrogen peroxide and lipid peroxides [28]. Through the action GST family of enzymes, GSH conjugates to a variety of electrophilic endogenous compounds and xenobiotics resulting in their efficient elimination from the body [29]. GSH and its oxidized form, glutathione disulfide (GSSG), form an important redox buffer that maintains cellular redox homeostasis [14]. As such, a decrease in the cellular GSH/GSSG ratio is often used as a marker of oxidative stress [30].

GSH is a ubiquitous tripeptide composed of glutamate, cysteine, and glycine. The rate-limiting enzymatic reaction in GSH biosynthesis is catalyzed by glutamate-cysteine ligase (GCL), a heterodimer comprising a catalytic (GCLC) and a modifier (GCLM) subunit. GCLC possesses all the catalytic activity of GCL, whereas GCLM serves to optimize the catalytic properties of GCL holoenzyme [31]. Our laboratories and others have developed several mouse models of GSH deficiency through genetic manipulation of the Gclc or Gclm genes [32]. The global Gclm-knockout (Gclm^KO) mice show no overt phenotype despite having 10–40% of normal tissue GSH levels [33]. The generally good health of Gclm^KO mice makes them a valuable model for studying the impact of chronic GSH deficiency. In the context of liver toxicity, Gclm^KO mice show high susceptibility to acute liver injuries caused by the drug acetaminophen (APAP) [34] and the environmental toxicant 2,3,7,8-tetrachlorodibenzodioxin (TCDD) [35]. To date, no liver cancer studies have been performed in Gclm^KO mice.

As an important component of hepatic antioxidant defense systems, the role of GSH in DX liver cytotoxicity and genotoxicity was investigated using the B6 Gclm^KO mouse model (Table 1) [13]. Male Gclm^KO mice and their wild-type (WT) littermates were exposed to a high dose DX orally for one or twelve weeks. Subchronic (12 weeks) DX exposure at 5000 ppm in drinking water caused only mild and yet comparable liver cytotoxicity in WT and Gclm^KO mice. A multitude of molecular changes in the liver were induced by high dose DX, including a transient compensatory NRF2 antioxidant response, persistent induction of cytochrome P450 2E1 (CYP2E1) enzyme and oxidative stress, elevated 8-OHdG levels and an increase in the γH2AX/H2AX ratio (a marker of DNA damage repair). Most importantly, Gclm^KO mice manifested increased sensitivity to DX-induced oxidative stress and genotoxicity in the liver. Collectively, these results provide molecular evidence linking redox dysregulation to DX liver genotoxicity, implying oxidative stress may act as a candidate MOA of DX liver carcinogenicity.

CYP2E1 plays a dominant role in DX metabolism and liver cytotoxicity, but contributes partially to DX liver genotoxicity

Two important observations from our Gclm^KO mice study pertain to the liver CYP2E1 enzyme. First, the hepatic oxidative stress resulting from subchronic DX exposure appeared to be associated with the progressive induction of CYP2E1 protein and activity [13]. This is consistent with the notion that, through one-electron reduction of oxygen, CYP2E1 activation leads to nonspecific formation of ROS, such as superoxide anion, hydrogen peroxide, hydroxyl radicals and lipid hydroperoxides [36]. These free radicals are potential inducers of cytotoxicity and mitogenic responses [37,38]; some of them, such as lipid peroxidation by-products malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE), are highly reactive and can form DNA-adducts leading to genotoxicity [39–41]. Second, there is a good correlation between hepatic CYP2E1 activity and plasma concentrations of β-hydroxyethoxyacetic acid (HEAA), the end-product of DX metabolism [13]. This observation suggests that liver CYP2E1 may be an important contributor to DX metabolism in vivo. To date, the metabolic pathways and the pathophysiological significance of DX metabolism in the liver are poorly understood. Therefore, our findings prompted the question whether CYP2E1 induction is essential to DX liver cytotoxicity and genotoxicity.

To address this research question, we replicated the high dose DX exposure study in male B6 global Cyp2e1-knockout (Cyp2e1^KO) mice (manuscript in preparation) (Table 1). The Cyp2e1^KO mouse model has been used in numerous toxicity studies and these mice display overall protection against liver injuries induced by drugs, xenobiotics or alcohol [36,42–44]. Such protection is believed to be attributable to abolished production of CYP2E1-derived toxic free radicals [42,45,46]. In our study, exposure to 5000 ppm DX in the drinking water for twelve weeks failed to cause liver cytotoxicity in Cyp2e1^KO mice; this was accompanied by blunted induction of hepatic oxidative stress and NRF2 antioxidant response. The plasma level of HEAA in DX-exposed Cyp2e1^KO mice was about 15% of that detected in DX-exposed WT mice. Unexpectedly, subchronic exposure to high dose DX induced a trend towards an increase in oxidative DNA damage (8-OHdG levels) and suppression of DNA damage repair (γH2AX/H2AX ratio) in the liver of Cyp2e1^KO mice. Collectively, these findings are indicative of a dominant role of CYP2E1 in DX metabolism and that CYP2E1 is responsible for DX-induced liver cytotoxicity and oxidative stress. However, our data also suggest that high dose DX-induced liver genotoxicity may, at least in part, involve CYP2E1-independent metabolic and molecular pathways.

Public health implications

DX is classified as a Group 2B human carcinogen by the International Agency for Research on Cancer (IARC) based on sufficient evidence in animal studies and inadequate human data [47,48]. Further, it has been recently classified as a substance of very high concern by the European Chemical Agency [49]. Drinking water is the dominant route of human exposure to DX affecting large numbers of individuals worldwide; such exposure is potentially extensive for people living near industrial release sites [4,6]. The monitoring data by US Environmental Protection Agency (EPA) under the Third Unregulated Contaminant Monitoring Rule (UCMR3) showed that 21% of 4864 tested US public water systems (PWS) contained detectable levels of DX (ranging from 0.07 to 33 μg/L) and 6.9% PWS exceeded the EPA health-based reference concentration of 0.35 μg/L [50]. This concerning widespread of DX contamination in the drinking water and the public health impact of such exposure have not been adequately addressed via regulatory standard setting, which is in large due to undetermined carcinogenic mechanisms of DX and the potential for low dose effects. In this context, of cautionary note, the genotoxic and cancer effective doses of DX used in ours and other animal studies (Table 1) are approximately 4,000- to 10,000-fold above the exposure levels reported in humans. However, these high dose studies are important and necessary to initially explore the adverse biological outcomes in a limited number of tested animals. It is therefore important, in future studies, to track the dose response for mechanistically relevant events identified in these studies to the lower dose range of environmental relevance. In addition, the plausible mechanistic roles of oxidative stress and CYP2E1 activation in DX genotoxicity are of high public health relevance, given that co-occurring contaminants (e.g., TCE and 1,1-dichloroethane (1,1-DCA)) and some dietary components (e.g., alcohol) may share a complimentary MOA to cause liver cancer [51–53] and therefore may greatly enhance the carcinogenic potency of the mixture.

Concluding remarks

In this mini-review, we present a brief discussion on a series of hypothesis-generating and following-up mechanistic studies that investigated redox changes associated with DX cytotoxicity and genotoxicity. These studies provide experimental evidence indicating that, following high dose DX exposure, (i) in the absence of liver cytotoxicity, DNA damage and alterations of oxidative stress defense systems are consistent and early molecular changes, (ii) deficiency in antioxidant glutathione increases the susceptibility to DX-induced oxidative DNA damage and DNA damage repair, and (iii) CYP2E1 plays a dominant role in DX metabolism and DX-induced hepatic oxidative stress and liver cytotoxicity, whereas CYP2E1 contributes partially to DX liver genotoxicity. The overall conclusions from these animal model studies support a direct genotoxic effect by high dose DX and imply that redox dysregulation and oxidative stress involving CYP2E1 activation may be causally involved in DX liver genotoxicity and potentially carcinogenicity. These mechanistic data derived from high dose studies can serve as important references to refine the assessment of carcinogenic pathways that may be triggered at environmentally relevant low doses in future human and animal studies.