Long-term Coexposure to Hexavalent Chromium and B[a]P Causes Tissue-Specific Differential Biological Effects in Liver and Gastrointestinal Tract of Mice

Abstract

Complex mixtures of environmental agents often cause mixture-specific health effects that cannot be accounted for by a single mechanism. To study the biological effects of exposure to a mixture of chromium-VI and benzo[a]pyrene (B[a]P), often found together in the environment, we exposed mice for 60 days to 0, 55, 550, or 5500 ppb Cr(VI) in drinking water followed by 90 days of coexposure to B[a]P at 0, 1.25, 12.5, or 125 mg/kg/day and examined liver and gastrointestinal (GI) tract for exposure effects. In the liver, the mixture caused more significant histopathology than expected from the sum of effects of the individual components, while in the GI tract, Cr(VI) alone caused significant enterocyte hypertrophy and increases in cell proliferation and DNA damage that were also observed in mice coexposed to B[a]P. Expression of genes involved in drug metabolism, tumor suppression, oxidative stress, and inflammation was altered in mixed exposures relative to control and to singly exposed mice. Drug metabolism and oxidative stress genes were upregulated and tumor suppressor and inflammation genes downregulated in the proximal GI tract, whereas most markers were upregulated in the distal GI tract and downregulated in the liver. Oral exposure to Cr(VI) and B[a]P mixtures appears to have tissue-specific differential consequences in liver and GI tract that cannot be predicted from the effects of each individual toxicant. Tissue specificity may be particularly critical in cases of extended exposure to mixtures of these agents, as may happen in the occupational setting or in areas where drinking water contains elevated levels of Cr(VI).

Keywords: B[a]P; complex mixtures; gene expression; heavy metals; hexavalent chromium.

FIG. 1.

A, Representative H&E (10×) images of the villus and crypt of proximal (PSI) and distal (DSI) sections of the GI tract. B, Representative H&E (2.5x and 40x) image of an atypical hyperplasia presented in the GI tract with 125 mg/kg/day B[a]P.

FIG. 2.

Representative H&E (10×) images of the periportal and centrilobular regions of the liver.

FIG. 3.

Immunohistochemical determination of Ki67 expression. A, PSI; B, DSI; and C, liver. For each condition 3 fields in each of 12 micrographs from male mice and another 12 from female mice, were taken and quantified In each field, 250–300 cells were analyzed for PSI and DSI and 75–100 cells were analyzed for liver. Magnification was 40×. Data shown represents the mean ± SEM. The symbol (*) denotes p < .05 different from the control condition of mice gavaged with corn oil. The character (a) denotes p < .05 different from mice exposed to 0 ppb Cr(VI) and 125 mg/kg/day B[a]P. The symbol (#) denotes p < .05 different from mice treated with the same dose of Cr(VI) but not treated with B[a]P.

FIG. 4.

Immunohistochemical determination of phospho-γH2AX. All details as in the legend to Figure 3.

FIG. 5.

Gene expression heat map for A, PSI; B, DSI; and C, liver. mRNA expression was measured by qRT-PCR for drug metabolism genes (Cyp1a1 and Cyp1b1), oxidative stress genes (Hmox1, Nrf2, Gclc, Gclm, Gpx1, Gpx4, Gstp1, Gsta4, and Gstm5), tumor suppressor genes (Cdkn1a/p21, Cdkn1b/p27, Cdkn2a/p16, Cdkn2b/p15, Cdkn2c/p18, Cdkn2c/p19, Hras, and P73) and pro inflammatory genes (Tnfα and Il6). All 16 dose combinations are shown.

FIG. 6.

Western immunoblot analysis of CYP1A1 expression. A Representative immunoblot of CYP1A1 in the tissues of mice exposed to the 4 extreme conditions tested, ie, control, 0 ppb Cr(VI) + 0 mg/kg/day B[a]P; 0 ppb Cr(VI) + 125 mg/kg/day B[a]P; 5500 ppb Cr(VI) + 0 mg/kg/day B[a]P; and 5500 ppb Cr(VI) + 125 mg/kg/day B[a]P. Immunoblot band intensities were quantified and are shown in (B), (C), and (D) for PSI, DSI, and liver, respectively, normalized to β-actin. A total protein extract from control and 8-hour B[a]P-treated mouse hepatoma Hepa-1 cells were used as a negative and positive control, respectively. The values shown correspond to the mean ± SEM of the band intensity relative to the indicated bands. *p < .05; **p < .01; ***p < .001; ****p < 0.0001.

FIG. 7.

Western immunoblot analysis of GCLC and NRF2. Representative immunoblot of GCLC and NRF2 in the same tissues as in Figure 5. Immunoblot band intensities were quantified and are shown in (A), (B), and (C) for PSI, DSI, and liver, respectively, normalized to β-actin. As a positive control for GCLC, a total protein extract from mouse kidney was used. For NRF2, A total protein extract from control and 8-hr B[a]P-treated mouse hepatoma Hepa-1 cells were used as a negative and positive control, respectively. The values shown correspond to the mean ± SEM of the band intensity relative to the indicated bands. Statistical comparisons as in Figure 6.