CYP2E1 in 1,4-dioxane metabolism and liver toxicity: insights from CYP2E1 knockout mice study

Abstract

1,4-Dioxane (DX), an emerging water contaminant, is classified as a Group 2B liver carcinogen based on animal studies. Understanding of the mechanisms of action of DX liver carcinogenicity is important for the risk assessment and control of this environmental pollution. Previous studies demonstrate that high-dose DX exposure in mice through drinking water for up to 3 months caused liver mild cytotoxicity and oxidative DNA damage, a process correlating with hepatic CYP2E1 induction and elevated oxidative stress. To access the role of CYP2E1 in DX metabolism and liver toxicity, in the current study, male and female Cyp2e1-null mice were exposed to DX in drinking water (5000 ppm) for 1 week or 3 months. DX metabolism, redox and molecular investigations were subsequently performed on male Cyp2e1-null mice for cross-study comparisons to similarly treated male wildtype (WT) and glutathione (GSH)-deficient Gclm-null mice. Our results show that Cyp2e1-null mice of both genders were resistant to DX-induced hepatocellular cytotoxicity. In male Cyp2e1-null mice exposed to DX for 3 months, firstly, DX metabolism to β-hydroxyethoxyacetic acid was reduced to ~ 36% of WT levels; secondly, DX-induced hepatic redox dysregulation (lipid peroxidation, GSH oxidation, and activation of NRF2 antioxidant response) was substantially attenuated; thirdly, liver oxidative DNA damage was at a comparable level to DX-exposed WT mice, accompanied by suppression of DNA damage repair response; lastly, no aberrant proliferative or preneoplastic lesions were noted in DX-exposed livers. Overall, this study reveals, for the first time, that CYP2E1 is the main enzyme for DX metabolism at high dose and a primary contributor to DX-induced liver oxidative stress and associated cytotoxicity. High dose DX-induced genotoxicity may occur via CYP2E1-independent pathway(s), potentially involving impaired DNA damage repair.

Keywords: Cytochrome p450; Liver carcinogenesis; Oxidative DNA damage; Oxidative stress; Water contaminant.

Fig. 1.. 1,4-Dioxane (DX) exposure regimen and liver histopathology.

(A) Scheme of DX exposure regimen. B6 male and female Cyp2e1^KO mice (12~14 weeks old; N = 6~8/group) were administered DX (5000 ppm) for 7 days (DX-7d) or 3 months (DX-3m) in their drinking water. Control (CON) mice received regular drinking water. ^#During the exposure course, two female Cyp2e1^KO mice from the DX-3m group died unexpectedly at 4^th and 11^th week; they were excluded from the study (N was reduced from 7 to 5 in this group). (B) Representative images of Hematoxylin and Eosin (H&E) stained liver sections. Minimal liver pathologies were noted in DX-exposed mice, manifesting sporadic hepatocyte death (blue arrows), inflammatory infiltration (red arrows) and stuffed stellate cells (yellow arrows). CV, central vein; PT, portal triad. (C) Liver histopathology score. Liver injury was evaluated based on the extent of hepatocyte injury, inflammation, reactive changes, and steatosis. Results are presented in Scatter plots with the group median at the line; steatosis was not observed in any of examined mice. *P < 0.05, **P < 0.01, by non-parametric Kruskal-Wallis test followed by Dunn’s multiple comparisons test.

Fig. 2.. Q-PCR analysis of gene expression in liver tissues of male mice.

(A) Gene expression in liver tissues of Cyp2e1^KO male mice by treatment groups relative to WT-CON levels. Examined genes included those involved in GSH metabolism, NRF2 antioxidant response, AHR signature target, and proinflammatory cytokines. The baseline level (dotted line at the value of 1) represented mRNA abundance in the WT-CON group after normalization to housekeeping genes (18s and Rplp0). Relative mRNA abundance is expressed as fold of control (WT-CON). Results are presented in Box plots showing individual data points (N = 5–7/group). ^aP < 0.05 when compared to WT-CON, ^bP < 0.05 when compared to Cyp2e1^KO-CON, by one-way ANOVA followed by Bonferroni multiple comparisons test. (B) Heatmap of differentially expressed genes among male mice of three genotypes by treatment groups. WT and Gclm^KO samples were derived from the previous study (Chen et al. 2022) and analyzed concurrently. Values represented the mean relative mRNA abundance (fold of WT-CON) of corresponding genes. CON, fed regular water; DX-7d, fed DX containing water for 1 week; DX-3m, fed DX containing water for 3 months.

Fig. 3.. Expression of redox sensitive proteins and glutathione concentrations in male liver tissues.

(A) Representative Western blotting images for protein expression using liver whole-cell lysates. One treatment-matched pooled samples (containing equal aliquots of 5–6 biological samples per group) derived from WT and Gclm^KO mice from the previous study (Chen et al. 2022) were analyzed concurrently. (B) Protein levels were quantitated by densitometry analysis and normalized to total protein loading per sample. Relative protein abundance was reported as fold of control (WT-CON). Group differences among Cyp2e1^KO mice were analyzed using one-way ANOVA followed by Bonferroni multiple comparisons test. Results are presented as mean ± S.D. (N = 6/group). *P < 0.05, **P < 0.01. (C) Liver concentrations of reduced (GSH) and oxidized glutathione (GSSG). Three treatment-matched pooled samples (each containing equal aliquots of two biological samples per group) derived from WT mice from the previous study (Chen et al., 2022) were analyzed concurrently. Data are reported as nmol/mg protein. Results are presented as mean ± S.D. (N = 3–6/group). Group differences were analyzed by two-way ANOVA followed by Bonferroni multiple comparisons test. *P < 0.05, **P < 0.01. CON, fed regular water; DX-7d, fed DX containing water for 1 week; DX-3m, fed DX containing water for 3 months.

Fig. 4.. DX-induced liver CYP2E1 induction and associated lipid peroxidation, circulating HEAA, and expression of two candidate DX metabolizing genes in male mice.

(A) Immunohistochemical staining for CYP2E1 and lipid peroxidation by-product 4-HNE in liver tissues. Magnification: 100x. (B) Semi-quantitation of plasma HEAA levels by untargeted metabolomic analysis. Data are reported as the area under the curve (AUC) of annotated HEAA peak after normalization. Results are presented as mean ± S.D. (N = 6/group). Group differences among DX-exposed mice were analyzed using one-way ANOVA followed by Bonferroni multiple comparisons test. ****P < 0.0001. (C) Q-PCR analysis of Cyp2b10 and Cyp2c29 genes. Relative mRNA abundance is expressed as fold of control (WT-CON) after normalization to housekeeping genes. Results are presented in Box plots showing individual data points (N = 5–7/group). Group differences were analyzed by two-way ANOVA followed by Bonferroni multiple comparisons test. ^a:P < 0.05, compared to treatment-matched WT mice; ^b:P < 0.05, compared to genotype-matched CON mice; ^c:P < 0.05, compared to genotype-matched DX-7d mice. CON, fed regular water; DX-7d, fed DX containing water for 1 week; DX-3m, fed DX containing water for 3 months. For each assay, treatment-matched samples derived from WT and Gclm^KO male mice from the previous study (Chen et al., 2022) were analyzed concurrently.

Fig. 5.. Oxidative DNA damage and DNA damage repair response in male liver tissues.

(A) DNA damage was assessed by measuring 8-OHdG in liver tissues using a competitive ELISA assay. Three treatment-matched pooled samples (each containing equal aliquots of two biological samples per group) derived from WT and Gclm^KO male mice from the previous study (Chen et al., 2022) were analyzed concurrently. Results are presented as mean ± S.D. (N = 3–6/group). Group differences were analyzed using two-way ANOVA followed by Bonferroni multiple comparisons test. P values of selected pair-wise comparisons were shown. P < 0.05 is considered statistical significance. (B-C) Western blotting analysis of DNA damage repair response marker (H2AX and γ-H2AX) and regulators (p53 and p21). One treatment-matched pooled samples (containing equal aliquots of 5–6 biological samples per group) derived from WT and Gclm^KO mice from the previous study (Chen et al. 2022) were analyzed concurrently. Representative Western blotting images (B) and quantitation of protein levels (C) by densitometry analysis. Relative protein abundance was reported as fold of control (WT-CON) after normalization to total protein loading per sample. Differences between treatment groups among Cyp2e1^KO male mice were analyzed using unpaired Student’s t-test. Results are presented as mean ± S.D. (N = 6/group). *P < 0.05. CON, fed regular water; DX-3m, fed DX containing water for 3 months.

Fig. 6.. Expression of proliferative and preneoplastic markers in male liver tissues.

(A-B) Western blotting analysis of marker proteins for cellular proliferation (PCNA) and preneoplastic lesion (GSTpi and CK7). One treatment-matched pooled samples (containing equal aliquots of 5–6 biological samples per group) derived from WT and Gclm^KO mice from the previous study (Chen et al. 2022) were analyzed concurrently. Representative Western blotting images (A) and quantitation of protein levels (B) by densitometry analysis. Relative protein abundance was reported as fold of control (WT-CON) after normalization to total protein loading per sample. Differences between treatment groups among Cyp2e1^KO male mice were analyzed using unpaired Student’s t-test. Results are presented as mean ± S.D. (N = 4–5/group). *P < 0.05. (C) Immunohistochemical staining for preneoplastic markers GSTpi and CK7 in liver tissues. Magnifications: 100x (400x inlets). CON, fed regular water; DX-3m, fed DX containing water for 3 months.

Fig. 7.. Proposed mechanism of action (MOA) of DX liver toxicity and carcinogenicity (associated with high dose and prolonged exposure).

The graphic scheme outlines the proposed MOA, which contains both a genotoxic (tumor initiation) and non-genotoxic (tumor promotion) pathways. Explicitly, DX oxidative metabolism to HEAA is mediated by CYP2E1 (~64%) and likely other CYP2 enzymes. In the genotoxic pathway (black outlines and lines), persistent induction of CYP2E1 mediates oxidative stress involving metabolism of DX and other endogenous and exogenous substrates; this process leads to cytotoxicity and oxidative DNA damage and activates NRF2 antioxidant defense as cellular adaptive response to redox dysregulation. In the non-genotoxic pathway (red outlines and lines), DX itself may act as a direct mitogen and inhibit DNA repair, which promotes cellular proliferation and genomic instability. The proposed MOA does not involve cytotoxicity in DX liver tumorigenicity.