Karen E. Wetterhahn

ARSENIC INDUCES OXIDANT STRESS AND NF-KB ACTIVATION IN CULTURED AORTIC ENDOTHELIAL CELLS

Chronic exposure to low levels of environmentally derived arsenite are associated with vascular diseases, such as arteriosclerosis. However, the cellular and molecular mechanisms for vascular disease in response to arsenic are not known. These studies investigated the hypothesis that nonlethal levels of arsenic increase intracellular oxidant levels, promote nuclear translocation of trans-acting factors, and are mitogenic. Incubation of second passage vascular endothelial cells with less than 5 microM arsenite for 4 h increased incorporation of [3H]thymidine into genomic DNA, while higher concentrations failed to stimulate or inhibit DNA synthesis. Within 1 h following addition of noncytotoxic concentrations of arsenite, oxidants accumulated and thiol status increased. During this time period, there was increased nuclear retention of NF-kappa B binding proteins and nuclear translocation of NF-kappa B also occurred in response to 100 microM H2O2. Supershift analysis demonstrated that p65/p50 heterodimers accounted for the majority of proteins binding consensus kappa B sequences in cells treated with arsenite or oxidants. The antioxidants, N-acetylcysteine or dimethylfumaric acid, increased intracellular thiol status and prevented both oxidant formation and translocation of NF-kappa B binding proteins in response to arsenite. These data suggest that arsenite initiates vascular dysfunction by activating oxidant-sensitive endothelial cell signaling.

Chromium(III) picolinate produces chromosome damage in Chinese hamster ovary cells.

Chromium(III) complexes currently being sold as dietary supplements were tested for their ability to cause chromosomal aberrations in Chinese hamster ovary cells. Complexes were tested in soluble and particulate forms. Chromium picolinate was found to produce chromosome damage 3‐fold to 18‐fold above control levels for soluble doses of 0.050, 0.10, 0.50, and 1.0 mH after 24 h treatment. Particulate chromium picolinate doses of 8.0 μg/cm2 (corresponding to a 0.10 mM solublized dose) and 40 μcm2 (0.50 mM) produced aberrations 4‐fold and 16‐fold above control levels, respectively. Toxicity was measured as a decrease in plating efficiency relative to controls. The above treatments produced e 86% survival for all doses except 1.0 mM chromium picolinate, which produced 69 ± 10% survival. Chromium nicotinate, nicotinic acid, and chromium(III) chloride hexahydrate did not produce chromosome damage at equivalent nontoxic doses. Damage was inferred to be caused by the picolinate ligand because picolinic acid in the absence of chromium was clastogenic. Data are evaluated in terms of their relevance to human exposure based on pharmacokinetic modeling of tissue accumulation and are discussed in terms of literature reporting toxic effects of picolinic acid.—Stearns, D. M., Wise, J. P., Sr., Patierno, S. R., Wetterhahn, K. E. Chromium(III) picolinate produces chromosome damage in Chinese hamster ovary cells. FASEB J. 9, 1643‐1648 (1995)

The interaction of chromium with nucleic acids

Native and denatured calf thymus DNA, and homopolyribonucleotides were compared with respect to chromium and protein binding after an in vitro incubation with rat liver microsomes, NADPH, and chromium (VI) or chromium (III). A significant amount of chromium bound to DNA when chromium (VI) was incubated with the native or the denatured form of DNA in the presence of microsomes and NADPH. For both native and denatured DNA the amount of protein bound to DNA increased with the amount of chromium bound to DNA. Denatured DNA had much higher amounts of chromium and protein bound than native DNA. There was no interaction between chromium(VI) and either form of DNA in the absence of the complete microsomal reducing system. The binding of chromium(III) to native or denatured DNA was small and relatively unaffected by the presence of microsomes and NADPH. The binding of chromium and protein to polyriboadenylic acid (poly(A], polyribocytidylic acid (poly(C], polyriboguanylic acid (poly(G] and polyribouridylic acid (poly(U] was determined after incubation with chromium(VI) in the presence of microsomes and NADPH. The magnitude of chromium and protein binding to the ribopolymers was found to be poly(G) much greater than poly(A) approximately equal to poly(C) approximately equal to poly(U). These results suggest that the metabolism of chromium(VI) is necessary in order for chromium to interact significantly with nucleic acids. The metabolically-produced chromium preferentially binds to the base guanine and results in DNA-protein cross-links. These findings are discussed with respect to the proposed scheme for the carcinogenicity of chromium(VI).

A prediction of chromium(III) accumulation in humans from chromium dietary supplements.

It has been proposed that 90% of Americans diets are deficient in the trace essential mineral chromium. Several chromium(III) dietary supplements are currently available to alleviate this deficiency. We show here that the same pharmacokinetic models that have been used to quantitate absorption of chromium(III) in humans predict that ingested chromium(III) will accumulate and be retained in human tissues for extended periods. Calculations were carried out with the popular supplement chromium picolinate as an example, but may be applied to any chromium(III) complex. Results from these calculations were compared to clinical data obtained from chromium(III) absorption/retention studies in humans. The models predict that chromium(III) can accumulate in human tissues to reach the levels at which DNA damage has been observed in animals and in vitro. The use of chromium supplements for extended periods or in excess dosages should be reevaluated in terms of these established models because the possible long‐term biological effects of chromium accumulation in humans are poorly understood.—Stearns, D. M., BelBruno, J. J., Wetterhahn, K. E. A prediction of chromium(III) accumulation in humans from chromium dietary supplements. FASEB J. 9, 1650‐1657 (1995)

Heme regulates hepatic 5-aminolevulinate synthase mRNA expression by decreasing mRNA half-life and not by altering its rate of transcription.

Hepatic 5-aminolevulinate (ALA) synthase, the first and rate-limiting enzyme in the heme biosynthetic pathway, is known to be feedback repressed by the end product of the pathway, heme. We investigated whether heme regulates ALA synthase mRNA expression transcriptionally or post-transcriptionally in primary cultures of chick embryo hepatocytes. 2-Propyl-2-isopropylacetamide increased the rate of transcription of the ALA synthase gene, whereas heme or an inhibitor of heme biosynthesis, desferrioximine, had no effect on the drug-induced transcription rate. Heme decreased the half-life of ALA synthase mRNA from approximately 3.5 h to 1.2 as recently reported by Drew and Ades (1989, Biochem. Biophys. Res. Commun. 162, 102-107). We also found that the heme-mediated decrease in mRNA stability was prevented by cycloheximide treatment, suggesting that the heme effect was mediated by a labile protein. These results support a model for hepatic ALA synthase regulation in which inducing drugs directly stimulate ALA synthase gene transcription, whereas heme regulates ALA synthase expression post-transcriptionally by modulating mRNA stability as well as by blocking translocation of ALA synthase enzyme into the mitochondrion.

Role of chromium(V), glutathione thiyl radical and hydroxyl radical intermediates in Chromium(VI)‐induced DNA Damage

Abstract In vitro chromium(VI) is unreactive toward DNA under physiological conditions. Therefore, the ability of chromium(VI) to damage DNA depends on the presence of cellular components capable of forming “reactive intermediates”; upon reaction with chromium(VI). We have examined the role of glutathione and hydrogen peroxide in chromium(VI)‐induced DNA damage in vitro. Reaction of glutathione with chromium(VI) produced significant levels of two chromium(V) complexes and glutathione thiyl radical, whereas reaction of chromium(VI) with hydrogen peroxide produced hydroxyl radical without producing detectable levels of chromium(V). Reaction of DNA with chromium(VI) in the presence of glutathione resulted in Cr‐DNA adducts with little DNA strand breakage. Reaction of DNA with chromium(Vl) in the presence of hydrogen peroxide produced the 8‐hydroxydeoxyguanosine adduct and extensive DNA strand breakage in the absence of significant Cr‐DNA adduct formation. These results suggest that the nature of chromium(VI)...

Mechanism of chromium(VI) carcinogenesis. Reactive intermediates and effect on gene expression.

Since chromium(VI) is unreactive toward DNA under physiological conditions in vitro, the ability of carcinogenic chromium(VI) compounds to damage DNA depends on the presence of cellular redox components that reduce chromium(VI) to reactive species capable of interacting with DNA. We have examined the role of glutathione and hydrogen peroxide in chromium(VI)-induced DNA damage in vitro. Upon reaction with chromium(VI), glutathione produced chromium(V) and glutathione thiyl radical reactive intermediates, whereas hydrogen peroxide produced chromium(V) and hydroxyl radical. Reaction of DNA with chromium(VI) in the presence of glutathione resulted in binding of chromium and glutathione to DNA with little or no DNA strand breakage. Reaction of DNA with chromium(VI) in the presence of hydrogen peroxide produced the 8-hydroxydeoxy-guanosine adduct and extensive DNA strand breakage in the absence of significant Cr-DNA adduct formation. These results suggest that the nature of chromium(VI)-induced DNA damage will be strongly dependent on reactive intermediates such as chromium(V), glutathione thiyl radical, and hydroxyl radical, produced by cellular components active in chromium(VI) metabolism. In order to assess the ability of chromium(VI)-induced DNA damage to affect the normal template function of DNA, we investigated the effects of chromium(VI) on steady-state mRNA levels of various genes in chick embryo liver in vivo, and compared the effects to the levels of DNA damage observed. Chromium(VI) induced DNA-protein and DNA interstrand cross-links in chick embryo liver in vivo and suppressed the induction of 5-aminolevulinic acid synthase and cytochrome P-450 mRNA expression by porphyrinogenic drugs. In contrast, chromium(VI) increased the basal levels of expression of these two inducible genes, but had little or no effect on the expression of the constitutive albumin, β-actin, and conalbumin genes. Comparison of the time course of formation and repair of DNA damage with that of changes in gene expression suggests that chromium(VI) may form a mono-adduct prior to formation of DNA cross-links, and that chromium(VI)-induced DNA lesions may target certain classes of genes and lead to changes in their expression.

Interlaboratory validation of a new assay for DNA-protein crosslinks

In 1992, a simple and sensitive assay for detecting DNA-protein crosslinks was developed [1]. In an effort to facilitate the greater use of the assay, a number of studies were conducted to evaluate its reliability and reproducibility. During this work, the assay was used to assess whether various metals and other compounds could induce crosslinks in cultured human lymphocytes (Epstein-Barr virus-transformed Burkitts Lymphoma cell line). Potassium permanganate, mercury chloride, lead nitrate, magnesium perchlorate, aluminum chloride, and cadmium chloride did not induce DNA-protein crosslinks at either cytotoxic or non-cytotoxic levels. Copper sulfate, arsenic trioxide, and potassium chromate induced DNA-protein crosslinks only at cytotoxic concentrations. Acute lethality of the cells was assessed immediately after exposure to metals by trypan blue exclusion while long-term lethality was assessed by cell proliferation and trypan blue exclusion following an incubation period of 5 days after exposure to the metal compound. All metals exhibited more toxicity in the long-term lethality assay compared to the short-term assay. The cultured human lymphocytes treated with various doses of lead acetate, cadmium chloride, arsenic trioxide and copper sulfate, as well as cis-platinum and chromate, were sent to four different laboratories to compare the reliability and reproducibility of the DNA-protein crosslink assay. Depending on the chemical studied, there were quantitative differences in the results observed among the various laboratories using the assay. However, all laboratories generally showed that cis-platinum, chromate, arsenic trioxide and copper sulfate induced DNA-protein crosslinks at levels that produced acute cytotoxicity, whereas cadmium chloride and lead acetate did not.

Chromium(VI) Toxicity: Uptake, Reduction, and DNA Damage

Much recent data supports the “uptake-reduction” model explaining the carcinogenicity of chromium(VI) compounds and the lack of carcinogenicity of chromium(III) com pounds. Cr(VI) readily enters cells by diffusion through a nonspecific anion channel, whereas cells are relatively impermeable to Cr(III). Glutathione appears to facilitate Cr(VI) uptake by reducing Cr(VI) to Cr(III) after it enters the cell, presumably keeping intracellular Cr(VI) concentration low and allowing for further Cr(VI) uptake. Some other nonenzymatic factors, for example, ascorbate and riboflavin, as well as enzymes, such as cytochrome P-450, DT-diaphorase, and the mitochondrial electron transport chain complexes, are capable of reducing Cr(VI) in vitro, but their contribution in vivo is not clear. Cr(VI), once reduced intracellularly, produces various forms of DNA damage including DNA interstrand crosslinks, DNA-protein crosslinks, DNA strand breaks, and Cr-DNA adducts. The pathway of Cr(VI) metabolism in different tissues appears to influence the type of “reactive intermediates” produced, for example, Cr(V) and radical species, and thus the nature and extent of DNA damage. This DNA damage presumably accounts for observed functional changes in DNA replication and transcription which may be crucial to the carcinogenicity of chromium(VI) compounds.

Tissue-specific changes in glutathione and cysteine after buthionine sulfoximine treatment of rats and the potential for artifacts in thiol levels resulting from tissue preparation☆

L-Buthionine-S,R-sulfoximine (BSO), a potent inhibitor of gamma-glutamylcysteine synthetase, is commonly used as an experimental tool for the specific depletion of glutathione. Since cysteine is a key precursor for glutathione biosynthesis, we investigated the possibility that BSO might also affect the free cysteine pool in rat liver and kidney tissues in vivo. Male CD(SD)BR rats (150-200 g) were injected ip with various doses of BSO (0.25-4.0 mmol/kg), and glutathione and cysteine were measured in liver and kidney using HPLC with electrochemical detection and/or spectroscopic techniques. No hepatotoxicity or nephrotoxicity was observed at the highest BSO dose (4.0 mmol/kg) used. BSO caused the expected decreases of hepatic and renal glutathione at all doses, although glutathione depletion was more rapid, was achieved at a lower BSO dose, and was more sustained in kidney than in liver. Hepatic cysteine levels nearly doubled 20 min after BSO treatment (1.0 mmol/kg, ip), but were not significantly different from control at later time points. In contrast, renal cysteine was significantly depleted from 20 min to 25 hr postinjection with a time course closely paralleling that of renal glutathione depletion. These changes are discussed in the context of models for inter- and intraorgan transport of glutathione and cysteine. We also provide evidence that an artifact, most likely the gamma-glutamyltranspeptidase (GGT)-initiated breakdown of glutathione, leads to a rapid postmortem increase of cysteine levels in liver and particularly in kidney of rats. Simultaneous decreases in GSH levels can be demonstrated in kidney. This artifact needs to be minimized in toxicological studies of glutathione and cysteine in kidney and other GGT-rich organs, as the measured levels of these thiols may not reflect the true concentrations occurring in vivo.