Cytotoxicity and oxidative mechanisms of different forms of chromium

Toxicology 180 (2002) 5
/22 www.elsevier.com/locate/toxicol

Cytotoxicity and oxidative mechanisms of different forms of chromium Debasis Bagchi a,*, Sidney J. Stohs a, Bernard W. Downs b, Manashi Bagchi b, Harry G. Preuss c a

Department of Pharmacy Sciences, Creighton University School of Pharmacy and Health Professions, 2500 California Plaza, Omaha, NE 68178, USA b InterHealth Research Center, Benicia, CA 94510, USA c Department of Physiology and Biophysics, Georgetown University Medical Center, Washington, DC 20007, USA

Abstract Chromium exists mostly in two valence states in nature: hexavalent chromium [chromium(VI)] and trivalent chromium [chromium(III)]. Chromium(VI) is commonly used in industrial chrome plating, welding, painting, metal finishes, steel manufacturing, alloy, cast iron and wood treatment, and is a proven toxin, mutagen and carcinogen. The mechanistic cytotoxicity of chromium(VI) is not completely understood, however, a large number of studies demonstrated that chromium(VI) induces oxidative stress, DNA damage, apoptotic cell death and altered gene expression. Conversely, chromium(III) is essential for proper insulin function and is required for normal protein, fat and carbohydrate metabolism, and is acknowledged as a dietary supplement. In this paper, comparative concentrationand time-dependent effects of chromium(VI) and chromium(III) were demonstrated on increased production of reactive oxygen species (ROS) and lipid peroxidation, enhanced excretion of urinary lipid metabolites, DNA fragmentation and apoptotic cell death in both in vitro and in vivo models. Chromium(VI) demonstrated significantly higher toxicity as compared with chromium(III). To evaluate the role of p53 gene, the dose-dependent effects of chromium(VI) were assessed in female C57BL/6Ntac and p53-deficient C57BL/6TSG p53 mice on enhanced production of ROS, lipid peroxidation and DNA fragmentation in hepatic and brain tissues. Chromium(VI) induced more pronounced oxidative damage in multiple target organs in p53 deficient mice. Comparative studies of chromium(III) picolinate and niacinbound chromium(III), two popular dietary supplements, reveal that chromium(III) picolinate produces significantly more oxidative stress and DNA damage. Studies have implicated the toxicity of chromium picolinate in renal impairment, skin blisters and pustules, anemia, hemolysis, tissue edema, liver dysfunction; neuronal cell injury, impaired cognitive, perceptual and motor activity; enhanced production of hydroxyl radicals, chromosomal aberration, depletion of antioxidant enzymes, and DNA damage. Recently, chromium picolinate has been shown to be mutagenic and picolinic acid moiety appears to be responsible as studies show that picolinic acid alone is clastogenic. Niacinbound chromium(III) has been demonstrated to be more bioavailable and efficacious and no toxicity has been reported. In summary, these studies demonstrate that a cascade of cellular events including oxidative stress, genomic DNA damage and modulation of apoptotic regulatory gene p53 are involved in chromium(VI)-induced toxicity and carcinogenesis. The safety of chromium(III) is largely dependent on the ligand, and adequate clinical studies are

* Corresponding author. Tel.: /1-402-280-2950; fax: /1-402-280-1883 E-mail address: [email protected] (D. Bagchi). 0300-483X/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 0 - 4 8 3 X ( 0 2 ) 0 0 3 7 8 - 5

6

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

warranted to demonstrate the safety and efficacy of chromium(III) for human consumption. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Chromium(VI); Chromium(III); In vitro; In vivo; Oxidative stress; DNA damage; Apoptotic cell death

1. Introduction Chromium occurs in the workplace predominantly in two valence states: hexavalent chromium [Chromium(VI)] and trivalent chromium [chromium(III)]. Hexavalent chromium compounds find extensive application in diverse industries, while trivalent chromium salts, including chromium chloride, niacin-bound chromium(III) or chromium polynicotinate, and chromium picolinate, are used as micronutrients and dietary supplements. Hexavalent chromium is widely known to cause allergic dermatitis, as well as toxic and carcinogenic effects in humans and animals (Norseth, 1981; Stohs and Bagchi, 1995; Kawanishi et al., 2002; von Burg and Liu, 1993). Hexavalent chromium-induced acute and chronic toxicity, neurotoxicity, dermatotoxicity, genotoxicity, carcinogenicity, immunotoxicity, and general environmental toxicity have been extensively demonstrated (von Burg and Liu, 1993; Barceloux, 1999). Soluble and insoluble hexavalent chromium salts have been demonstrated to induce morphological and neoplastic transformation and mutagenicity in murine and human cells (Patierno et al., 1988). Hexavalent chromium compounds induce dose-dependent cytotoxicity and anchorage independence in cultured human diploid fibroblasts (Biedermann and Landolph, 1987). Soluble hexavalent chromium compound calcium chromate (CaCrO4) has been shown to induce dose-dependent cytotoxicity in C3H/10T1/2 mouse embryo cells and dose-dependent cytotoxicity and mutagenesis in Chinese hamster ovary cells, while insoluble salt lead chromate (PbCrO4) induced cytotoxicity as well as morphological and neoplastic transformation in C3H/10T1/2 mouse embryo cells (Patierno et al., 1988). Shi and Dalal (1989, 1990) have used electron spin resonance (ESR) to demonstrate the formation of long-lived chromium(V) intermediates in

the reduction of chromium(VI) by glutathione reductase (GR) in the presence of NADPH, and the generation of noxious hydroxyl radicals. Hydrogen peroxide suppresses chromium(V) and enhances the formation of hydroxyl radicals through a chromium(V)-catalyzed Fenton-like reaction. Subsequent investigations with superoxide dismutase (SOD) showed no significant participation of superoxide anion in the production of hydroxyl radicals. Related studies by Jones et al. (1991) have provided evidence that suggests that hydroxyl radicals are generated from a chromium(V) intermediate that is responsible for causing DNA single strand breaks. Kawanishi et al. (1986) have demonstrated that chromium(VI) produces noxious ROS including superoxide anion, singlet oxygen and hydroxyl radicals, through the formation of chromium(V) intermediates. The mechanism of DNA cleavage by chromium(VI) in the presence of hydrogen peroxide was investigated. ROS were produced by the decomposition of Cr(V)(O2)43 ion, resulting in DNA damage. ESR was used for the detection and quantification of hydroxyl radicals. These results indicate that chromium(V) complexes produced in the reduction of chromium(VI) by cellular reductants react with hydrogen peroxide to generate hydroxyl radicals, which may be the initiators of the primary events in chromium(VI) cytotoxicity and carcinogenicity. It has been reported that trivalent chromium salts are poorly absorbed through the gastrointestinal tract. Hexavalent chromium compounds are approximately 1000-fold more cytotoxic and mutagenic than trivalent chromium compounds in cultured diploid human fibroblasts, but both hexavalent and trivalent chromium compounds induce dose-dependent anchorage independence in human diploid fibroblasts (Biedermann and Landolph, 1990; Levis et al., 1978). The chromate ion [CrO4] 2, the dominant form of chromium(VI) in neutral aqueous solutions, can readily cross cellu-

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

lar membranes via non-specific anion carriers (Danielsson et al., 1982), while chromium(III) is poorly transported across membranes. The differences in membrane transport may explain the differences in the abilities of these two valence states of chromium to induce the formation of ROS and produce oxidative tissue damage. This paper highlights the results of a series of in vitro and in vivo studies, which demonstrate the concentration- and time-dependent effects of chromium(VI) and chromium(III) salts on increased production of ROS including superoxide anion and hydroxyl radicals, enhanced excretion of urinary lipid metabolites, lipid peroxidation, DNA fragmentation and apoptotic cell death in both in vitro and in vivo models. The effects of acute oral doses of 0.10 LD50 and 0.50 LD50 of chromium(VI) were assessed in female C57BL/ 6Ntac and p53-deficient C57BL/6TSG p53 mice on enhanced production of superoxide anion, lipid peroxidation and DNA fragmentation in hepatic and brain tissues to evaluate the role of apoptotic regulatory p53 gene. Concentration-dependent comparative induction of oxidative stress including superoxide anion production, lipid peroxidation and DNA fragmentation were also assessed between chromium(III) picolinate and niacinbound chromium(III), two popular trivalent chromium supplements, in cultured J774A.1 murine macrophage cells.

2. Materials and methods 2.1. Chemicals Chromium(III) chloride hexahydrate and sodium dichromate [chromium(VI)], 2,3-dihydroxybenzoic acid, 2,5-dihydroxybenzoic acid, salicylic acid, and HPLC-grade water were purchased from Aldrich Chemical Co. (Milwaukee, WI). Sodium acetate trihydrate and citric acid used for HPLC analysis were purchased from Fluka (Buchs, Switzerland). All other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO), and were of analytical grade or higher and used without purification.

7

2.2. Cell cultures and treatment Human chronic myelogenous leukemic K562 cells and murine macrophage J774A.1 cells were obtained from the American Type Culture Collection (Rockville, MD). The human chronic myelogenous leukemic cells were maintained and grown in RPMI 1640 medium with 10% fetal calf serum, while the J774A.1 cell line was maintained and grown in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum. The cultured K562 cells (1 /106 cells/35 mm petri dish) were incubated with 0, 12.5 and 25 mM concentrations of chromium(VI) for 0, 24, or 48 h. Cultured J774A.1 cells were plated at 25 /104 cells/35 mm petri dish (Corning Inc., Corning, NY), and incubated with either 0, 0.20, 0.40 or 0.60 mM concentrations of chromium(VI) at 37 8C for 0, 24 and 48 h. Cell viability was determined by Trypan blue exclusion technique. 2.3. Animals and treatment All animals were housed in a controlled environment at 25 8C with a 12 h light and 12 h dark cycle, and acclimated for at least 3
/5 days before use. Female Sprague
/Dawley rats (160
/180 g) were purchased from Sasco, Inc. (Omaha, NE). Chromium(VI)-treated rats received a daily dose of 2.5 mg [0.05 LD50], 10 mg [0.20 LD50], or 25 mg [0.50 LD50] sodium dichromate/kg body weight orally in water, while chromium(III)-treated rats received a daily dose of 895 mg [0.50 LD50] chromium(III) chloride hexahydrate/kg body weight orally in water. For acute toxicity study, frozen urine samples were collected on dry ice on 0, 24, 48, 72 and 96 h post-treatment, while groups of control and treated animals were also sacrificed at 48 h post treatment to obtain peritoneal exudate cells (primarily macrophages) and hepatic mitochondria and microsomes. Peritoneal exudate cells (primarily macrophages) were isolated using Hepes buffer, pH 7.4 (140 mM NaCl, 5 mM KCl, 10 mM glucose, 2 mM Ca2 and 20 mM Hepes), which was injected into the peritoneal cavity, and subsequently aspirated, as demonstrated previously by us (Bagchi et al., 1995a,b). After withdrawl of the

8

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

peritoneal macrophages, livers were immediately removed and placed in ice-cold 50 mM Tris
/KCl buffer (pH 7.4) containing 150 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, and 10% glycerol. The livers were homogenized with 5 ml buffer per gm in a Potter
/Elvehjem homogenizer fitted with Teflon pestle (four 30-s strokes). Subcellular fractionation was achieved by differential centrifugation as previously described by us (Bagchi et al., 1995a,b). For chronic toxicity study, animals were given a daily oral dose up to 90 days, frozen urine samples were collected on dry ice on 0, 30, 60 and 90 days of treatment, and groups of control and treated animals were sacrificed on 0, 30, 60 and 90 days of treatment to obtain liver samples. Liver samples were processed to obtain hepatic mitochondria and microsomal membranes (Bagchi et al., 1995a,b). Identification and quantitation of urinary lipid metabolites including malondialdehyde (MDA), formaldehyde (FA), acetaldehyde (ACT) and acetone (ACON), were determined as previously demonstrated by us (Bagchi et al., 1995a,b). In summary, 2,4-dinitrophenylhydrazine derivatives of urinary lipid metabolites were quantitated by HPLC at 330 nm using an acetonitrile:water (49:51 v/v) mobile phase at a flow rate of 1 ml/min as described earlier (Bagchi et al., 1995a,b). Female C57BL/6Ntac and p53-deficient C57BL/ 6TSG p53 mice (4-weeks old) were obtained from Taconic (Germantown, NY). The mice were treated with single oral doses of 0, 19 mg [0.10 LD50] and 95 mg [0.50 LD50] sodium dichromate per kg body weight orally in water and sacrificed after 24 h. Hepatic and brain tissues were collected and preserved at /80 8C for biochemical assays.

production by peritoneal macrophages was also measured by the method of Prodczasy and Wei (1988), which is based on the reduction of iodonitrotetrazolium (INT) and as described previously by us (Bagchi et al., 1995a,b).

2.4. Superoxide anion production

Cell or frozen tissue samples were homogenized in lysis buffer (5 mM Tris
/HCl, 20 mM EDTA, 0.5% Triton X-100, pH 8.0). Homogenates were centrifuged at 27 000/g for 20 min to separate intact chromatin in the pellets from fragmented DNA in the supernatant fractions. Pellets were resuspended in 0.5 N perchloric acid, and 5.5 N perchloric acid was added to the supernatant fractions to reach a concentration of 0.5 N. Samples were heated at 90 8C for 15 min and centrifuged at 1500 /g for 10 min to remove

Superoxide anion production was determined on the basis of cytochrome c reduction following the method of Babior et al. (1973) and as previously described by us (Bagchi et al., 1995a). Absorbance values were converted to nmoles of cytochrome c reduced per 15 min/mg protein or nmoles cytochrome c reduced/15 min/number of cells used in the study, using the extinction coefficient 2.1 /104 per M/cm. Superoxide anion

2.5. Hydroxyl radical detection by HPLC The detection of hydroxyl radicals based on the formation of 2,3-dihydroxybenzoic and 2,5-dihydroxybenzoic acid was determined using HPLC equipped with a Waters 460 electrochemical detector as previously described by us (Bagchi et al., 2000) using 100 mM sodium salicylate as the substrate. The hydroxylated products of salicylic acid after interaction with hydroxyl radicals, 2,3dihydroxy- and 2,5-dihydroxybenzoic acids, were eluted with a mobile phase containing 0.03 mol/l sodium acetate and 0.03 mol/l of citric acid (pH 3.6) at a flow rate of 1 ml/min. The detection potential was maintained at /0.6 V, employing a glassy carbon working electrode and an Ag/AgCl reference electrode. Retention times for the peaks of 2,3-dihydroxy- and 2,5-dihydroxybenzoic acids were verified by injecting authentic standards. 2.6. Lipid peroxidation Lipid peroxidation was determined as previously described (Bagchi et al., 1995a,b), based on the formation of thiobarbituric acid reactive substances (TBARS). MDA was used as the standard. A molar extinction coefficient of 1.56 /105 per M/cm was used. 2.7. DNA fragmentation

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

protein. Resulting supernatant fractions were reacted with diphenylamine for 16
/20 h at room temperature. Absorbances were measured at 600 nm. DNA fragmentation in control samples is expressed as percent of total DNA appearing in the supernatant fraction. Treatment effects are reported as percent of control fragmentation (Bagchi et al., 2000). 2.8. Cell cycle analysis Chromium(VI)-treated and untreated cells were collected by centrifugation and resuspended in 1 ml of 70% ethanol, and cell cycle analyses were conducted as described by us previously. Briefly, cells were resuspended in Telford’s reagent containing EDTA, RNAase, propidium iodide, Triton X-100, and PBS, incubated at 4 8C for 24 h, and analyzed using a Becton Dickinson FACSTAR Plus Flow Cytometer (San Jose, CA) equipped with a CELL QUEST Software Program. The percentages of cells within the Go/G1, S, and G/ M phases of the cell cycle were determined by analysis with the software program. The percentages of apoptotic cells were calculated from the forward angle light scatter versus linear red fluorescence histogram of the Ao population as described by Telford et al. (1991) and by us (Bagchi et al., 2000). 2.9. Detection of apoptotic cells Cells were fixed in 4% buffered paraformaldehyde and embedded in paraffin. The terminal deoxynucleotidyl transferase (TdT)-mediated dUTP biotin nick end labeling (TUNEL) method, developed by Gavrieli et al. (1992) and by us (Bagchi et al., 1998), was used to visualize DNA fragmentation at the single cell level. In the TUNEL method, 3-OH DNA ends generated by DNA fragmentation are nick end labeled with biotinylated dUTP, introduced by TdT, and counter stained with propidium iodide. A MEBSTAIN Apoptosis Kit from Medical and Biological Laboratories (Nagoya, Japan) was used for fluorescent detection of the dead cells versus live cells.

9

2.10. MTT [3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide] cleavage assay Respiratory effects of chromium(VI) on exposed J774A.1 cells were assessed by determining the ability of the cells to reduce the tetrazolium dye MTT based on the activation of the enzyme succinate dehydrogenase as described by Mosmann (1983) and by us (Bagchi et al., 1998). The optical density was read at 560 nm and the amount of formazan produced was calculated by using molar extinction coefficient of 51 000. 2.11. Statistical methods The presence of significant differences between groups was determined using analysis of variance (ANOVA), with Scheffe’s S method as the post hoc test. Each value is the mean9/S.D. from four to six experiments. The level of statistical significance employed in all cases was P B/0.05.

3. Results 3.1. Enhanced excretion of urinary lipid metabolites The effects of an acute oral 0.50 LD50 dose of chromium(VI) and chromium(III) on urinary excretion of the lipid oxidation products MDA, FA, ACT, and ACON between 0 and 96 h after treatment are presented in Table 1. Time-dependent effects of both chromium(VI) and chromium(III) on the excretion of urinary lipid metabolites were assessed, and the maximal excretion of the four lipid metabolites was observed between 48 and 72 h. Approximately 1.5
/4.7-fold increases in the four lipid metabolites were observed. The greatest increase occurred in the excretion of MDA. Chromium(VI)-induced more dramatic excretion of all four metabolites as compared with chromium(III), which demonstrates that chromium(III) is much less toxic compared with chromium(VI). In two separate sets of experiments, we assessed the effect of chronic administration of 2.5 mg (0.05 LD50) and 10 mg (0.20 LD50) sodium dichromate/kg body weight/day over a period

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

10

Table 1 Urinary excretion of lipid metabolites (nmoles/kg body weight/6.0 h) after a single acute oral 0.50 LD50 dose of sodium dichromate [chromium(VI); 25 mg/kg body weight] or chromium(III) chloride hexahydrate [chromium(III); 895 mg/kg body weight] Hours post-treatment Metabolite

Treatment

0

24

48

72

96

MDA

Control Chromium(VI) Chromium(III) Control Chromium(VI) Chromium(III) Control Chromium(VI) Chromium(III) Control Chromium(VI) Chromium(III)

2.19/0.2
/
/ 50.1/4.8
/
/ 2.3/0.3
/
/ 4.1/0.3
/
/

2.2/0.3 3.2/0.4a 2.4/0.3 49.4/5.1 71.6/5.5a 54.3/8.2 2.3/0.2 5.5/0.6a 3.1/0.2 4.0/0.3 7.8/0.7a 5.1/0.6

2.3/0.2 9.4/0.8a 3.4/0.5 47.8/4.3 132.4/9.0a 90.8/12.2a 2.2/0.3 10.1/0.7a 5.1/0.4a 3.8/0.4 11.6/1.1a 9.5/1.2a

2.0/0.2 9.3/0.7a 3.3/0.4a 45.9/5.3 110.2/7.8a 86.6/9.5a 2.3/0.2 10.3/0.9a 5.6/0.8a 3.9/0.3 12.3/1.2a 8.3/0.7a

2.1/0.3 5.8/0.7a 2.3/0.2 47.6/4.7 90.7/8.7a 66.8/7.2a 2.2/0.3 7.8/0.8a 4.4/0.3a 3.8/0.3 10.3/0.7a 5.6/0.4a

FA

ACT

ACON

Female Spraque
/Dawley rats were treated with a single oral dose of sodium dichromate (25 mg/kg body weight) or chromium chloride hexahydrate (895 mg/kg body weight) in water. Control animals received the vehicle. Urine samples were collected for 6 h over dry ice. Metabolites were determined by HPLC. Each value is the mean9/S.D. of four animals. a P B/0.05 with respect to the corresponding control group.

of 0
/90 days (Tables 2 and 3). Following chronic administration of 2.5 mg sodium dichromate/kg body weight/day for 90 days, approximately 1.8
/2.2-fold increases of the four lipid metabolites were observed between 0 and 90 days of treatment (Table 2), while chronic admin-

istration of 10 mg sodium dichromate/kg body weight/day for 90 days resulted in approximately 2.6
/4.1-fold increases of the four lipid metabolites. These two experiments demonstrate the low dose chronic toxicity of hexavalent chromium in vivo.

Table 2 Urinary excretion of lipid metabolites (nmoles/kg body weight/6.0 h) after daily treatment with 0.05 LD50 (2.5 mg/kg body weight/day) of sodium dichromate for 90 days Days of treatment Metabolite

Treatment

MDA

Control Chromium(VI) Control Chromium(VI) Control Chromium(VI) Control Chromium(VI)

FA ACT ACON

0 2.39/0.3 46.49/5.1 2.49/0.3 4.39/0.3 -

30

60

90

2.99/0.4 5.89/0.6a 49.39/5.4 66.49/8.1 2.69/0.2 5.89/0.6a 5.39/0.4 8.19/1.1a

4.29/0.5 9.39/1.2a 55.29/6.1 101.39/9.6a 3.79/0.5 8.09/0.9a 6.39/0.7 11.29/1.2a

4.59/0.5 9.89/1.3a 58.29/6.5 102.69/12.4a 5.09/0.6 8.89/0.7a 7.49/0.8 12.49/1.8a

Female Spraque
/Dawley rats were treated with 2.5 mg/kg body weight/day sodium dichromate orally in water. Urine samples were collected for 6 h over dry ice. Metabolites were quantitated by HPLC. Each value is the mean9/S.D. of four animals. a P B/0.05 with respect to the corresponding control group.

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

11

Table 3 Urinary excretion of lipid metabolites (nmoles/kg body weight/6.0 h) after daily treatment with 0.20 LD50 (10 mg/kg body weight/day) of sodium dichromate for 90 days Days of treatment Metabolite

Treatment

MDA

Control Chromium(VI) Control Chromium(VI) Control Chromium(VI) Control Chromium(VI)

FA ACT ACON

0 2.09/0.2 43.99/3.5 2.09/0.2 4.09/0.4 -

30

60

90

3.49/0.4 8.69/1.0a 49.19/3.8 91.59/8.8a 2.89/0.3 11.59/0.7a 5.09/0.5 11.49/1.4

4.19/0.6 12.29/0.8a 55.49/6.6 143.59/10.2a 3.69/0.4 10.99/0.9a 5.59/0.4 16.19/0.7a

4.89/0.6 11.89/1.3a 56.89/5.4 142.39/8.6a 4.89/0.4 8.99/0.9a 6.09/0.8 13.59/0.9a

Female Spraque
/Dawley rats were treated with 10 mg/kg body weight/day sodium dichromate orally in water. Urine samples were collected for 6 h over dry ice. Metabolites were quantitated by HPLC. Each value is the mean/S.D. of four animals. ¯ a P B/0.05 with respect to the corresponding control group.

3.2. Superoxide anion production Cytochrome c reduction was performed to assess superoxide anion production. The effects of chromium(VI) and chromium(III) on the enhanced production of superoxide anion by peritoneal macrophages are presented in Table 4. The data is presented as nanomoles of cytochrome c reduced produced/15 min. Chromium(III) administration (895 mg/kg body weight) increased the production of superoxide anion in vivo based on cytochrome c and INT reduction as compared with the peritoneal exudate cells from control animals by 1.7- and 1.5-fold, respectively, at 48 h post-treatment, while increases of 3.7- and 3.6fold were observed in cytochrome c and INT

reduction, respectively, following chromium(VI) administration (Table 4). Human chronic myelogenous leukemic K562 cells were treated with 12.5 and 25 mM concentrations of chromium(VI) for 24 h; approximately 2.7- and 4.8-fold increases in cytochrome c reduction were observed, respectively, (Table 5). Cultured J774A.1 murine macrophage cells were treated with 0, 10, 30 and 50 mg/ml concentrations of chromium(III) picolinate or chromium(III) nicotinate for 24 h. Approximately 1.0-, 1.1- and 1.5-fold increases in cytochrome c reduction were observed following incubation of these cells with 10, 30 and 50 mg/ml concentrations of chromium(III) picolinate, respectively, as compared with the control untreated cells. While under these same

Table 4 Production of superoxide anion by peritoneal macrophages and hepatic mitochondrial and microsomal lipid peroxidation (TBARS content) following treatment of rats with either chromium(III) chloride hexahydrate (895 mg/kg body weight) or sodium dichromate [Chromium(VI) (25 mg/kg body weight] Cytochrome c (nmol) reduced/3/106 macrophage cells/15 min

Control Chromium(VI) Chromium(III)

51.49/6.3 188.79/13.5a 84.79/10.2a

INT (nmol) reduced/3/106 macrophage cells/15 min

26.39/3.4 93.49/10.7a 38.99/5.2a

TBARS content (nmol/mg protein) Mitochondria

Microsomes

5.79/0.7 10.39/1.3a 6.99/0.8a

3.99/0.5 8.89/0.8a 5.39/0.6a

Animals were sacrificed 48 h post-treatment. Each value represents the mean9/S.D. of four to six animals. MDA was used as the standard. a P B/0.05 with respect to the control group.

Cultured K562 cells were treated with 12.5 or 25 mM concentration of sodium dichromate [chromium(VI)]. See Section 2 for details. Values with non-identical superscripts are significantly different (P B/0.05).

0.31a 3.31b 4.38c 0.139/0.02 2.519/0.53 3.499/0.66 0.189/0.05 0.809/0.21 0.899/0.14 99.09/6.7a 262.69/18.7b 479.69/41.6c Control Chromium(VI) 12.5 mM Chromium(VI) 25 mM

2,3-DHBA (nmoles/ml) 2,5-DHBA (nmoles/ml)

Cytochrome c reduction (nmoles reduced/ 15 min/1/106 cells)

Hydroxyl radical production

Total (nmoles/ml)

8.179/1.80a (100) 18.069/1.12b (221) 24.849/1.86c (304)

DNA fragmentation (% control)

0a 39.59/6.3b 0a

Apoptotic cell death (%)

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

Treatment

Table 5 Concentration-dependent changes in enhanced production of superoxide anion and hydroxyl radicals, changes in intracellular oxidized states, DNA fragmentation and apoptotic cell death in chronic myelogenous leukemic K562 cells following 24 h treatment with sodium dichromate [chromium(VI)]

12

conditions, approximately 1.0-, 1.0- and 1.2-fold increases in cytochrome c reduction were observed following treatment with the same concentrations of niacin-bound chromium(III) (Table 6). Cytochrome c reduction was assayed as a marker of superoxide anion production in the liver and brain tissues obtained from both C57BL/ 6Ntac and p53 deficient C57BL/6TSG p53 mice. Increases in cytochrome c reduction were observed in the liver and brain tissues of both strains of mice at both 0.10 LD50 and 0.50 LD50 doses. Following treatment with acute oral 0.10 LD50 and 0.50 LD50 doses of chromium(VI) approximately 2.5- and 6.0-fold increases in cytochrome c reduction were observed in the liver, and 2.5- and 4.1-fold increases in cytochrome c reduction in the brain tissues of C57BL/6Ntac mice, respectively, as compared with the control animals (Tables 7 and 8). While under these same conditions, chromium(VI)-induced 5.4- and 11.1-fold increases in cytochrome c reduction in the liver, 3.7- and 7.7fold increases in cytochrome c reduction in brain tissues of p53 deficient C57BL/6TSG p53 mice, respectively, as compared with the control untreated mice (Tables 7 and 8). Thus, significantly higher production of superoxide anion was observed for chromium(VI) in both the liver and brain tissues of p53 deficient C57BL/6TSG p53 mice as compared with the corresponding control animals. 3.3. Hydroxyl radical production Following treatment of the human chronic myelogenous leukemic cells for 24 h with 12.5 and 25 mM concentrations of chromium(VI), approximately 10.7- and 14.1-fold increases in hydroxyl radical production were observed, respectively, as compared with the untreated cells (Table 5). Cultured J774A.1 murine macrophage cells were treated with 0, 10, 30 and 50 mg/ml concentrations of chromium(III) picolinate or chromium(III) nicotinate for 24 h. Approximately 1.0-, 1.4- and 1.6-fold increases in hydroxyl radical production were observed following incubation of these cells with 10, 30 and 50 mg/ml concentrations of chromium(III) picolinate, respectively, as compared with the control untreated cells. While under

Cultured J774A.1 macrophage cells (25/104 cells/35 mm petri dish) in 2 ml of DMEM were incubated for 2 h to allow cell adherence and various concentrations of chromium picolinate or chromium nicotinate were added to the cultures. Assays were conducted after 24 h of incubation as described in the Section 2. Data are expressed as the mean values of four experiments9/S.D. Values with non-identical superscripts are significantly different (P B/0.05).

3.769/0.76a (100) 3.959/0.80a (105) 4.369/0.53a,b (116) 4.919/0.65b (131) 3.769/0.76a (100) 3.889/0.69a (103) 4.329/0.80a,b (115) 6.199/0.65b (165) 0.28a 0.27a 0.42b 0.42b 0 10 30 50

0.149/0.02 0.149/0.02 0.199/0.03 0.259/0.06

0.149/0.04 0.149/0.03 0.219/0.13 0.199/0.04

0.28a 0.28a 0.40b 0.44b

0.149/0.02 0.139/0.04 0.249/0.03 0.239/0.05 6.759/0.82a 6.969/0.55a 7.009/0.29a 8.359/0.79b 6.759/0.82a 6.499/108a 7.289/0.73a 10.119/0.82b

0.149/0.04 0.149/0.02 0.189/0.02 0.199/0.05

Niacin-bound chromium 2,3DHBA 2,5DHBA

2,3DHBA

Total 2,5DHBA

Chromium picolinate Niacin-bound chromium Chromium picolinate

Niacin-bound chromium

Total Chromium picolinate

DNA fragmentation (% Control) Formation of 2,3- and 2,5-dihydroxybenzoic acids (nmoles/25/104 cells) Cytochrome c reduction (nmoles cytochrome c reduction/15 min Concentration mg/ml

Table 6 Chromium(III) picolinate and niacin-bound chromium(III)-induced cytochrome c reduction, hydroxyl radical production and DNA fragmentation in cultured J774A.1 macrophage cells

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

13

these same conditions, approximately 1.0-, 1.5and 1.5-fold increases in hydroxyl radical production were observed following treatment with the same concentrations of niacin-bound chromium(III) (Table 6). 3.4. Lipid peroxidation The effects of chromium(VI) and chromium(III) on lipid peroxidation in hepatic mitochondria and microsomes based on the TBARS assay are summarized in Table 4. Mitochondrial fractions from control animals exhibited higher levels of lipid peroxidative activity than microsomal preparations. Increases in lipid peroxidation of 1.8and 2.2-fold occurred in hepatic mitochondria and microsomes, respectively, 48 h after the oral administration of 25 mg sodium dichromate(VI)/ kg body weight, while increases of 1.2- and 1.4fold, respectively, were observed after 895 mg chromium(III) chloride hexahydrate/kg body weight (Table 4). The chronic effects of chromium(VI) on lipid peroxidation in hepatic mitochondria and microsomes are summarized in Table 9. An age-dependent increase in lipid peroxidation was observed over the 90 days of this study. Lipid peroxidation increased 54 and 45% in mitochondria and microsomes, respectively, in control animals between days 0 and 90 (Table 9). In mitochrondrial fractions, increases in the TBARS content of 1.6, 2.4- and 2.1-fold were observed at 30, 60, and 90 days of treatment with chromium (VI), respectively, relative to control values. Increases in the TBARS content of about 1.4-, 1.9- and 2.0-fold were observed in hepatic microsomes at 30, 60, and 90 days of treatment, respectively, relative to control values (Table 9). Following treatment of the Sprague
/Dawley rats with 0.05 LD50 dose (2.5 mg/kg body weight/day) of chromium (VI) for 90 days approximately, 1.8-and 1.7-fold increases in TBARS content were observed in the hepatic mitochondria and microsomes, respectively, following treatment of the rats for 90 days with chromium(VI) (data not shown). Lipid peroxidation was assayed as a marker of oxidative damage to the lipids in the liver and brain tissues obtained from both C57BL/6Ntac

14

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

and p53 deficient C57BL/6TSG p53 mice. Increases in lipid peroxidation were observed in the liver and brain tissues of both strains of mice at both 0.10 LD50 and 0.50 LD50 doses. Following treatment with acute oral 0.10 LD50 and 0.50 LD50 doses of chromium(VI) approximately 1.4 and 3.3fold increases in lipid peroxidation were observed in the liver, and 1.7- and 3.5-fold increases in lipid peroxidation in the brain tissues of C57BL/6Ntac mice, respectively, as compared with the control animals (Tables 7 and 8). While under these same conditions, chromium(VI)-induced 1.8- and 10.8fold increases in lipid peroxidation in the liver, 2.4and 7.3-fold increases in lipid peroxidation in the brain tissues of p53 deficient C57BL/6TSG p53 mice, respectively, as compared with the control untreated mice (Tables 7 and 8). Thus, significantly higher lipid peroxidation was observed in both the liver and brain tissues of p53 deficient C57BL/6TSG p53 mice as compared with the corresponding control animals. 3.5. DNA fragmentation Following treatment of the human chronic myelogenous leukemic K562 cells with 12.5 and 25 mM concentrations of chromium(VI) for 24 h, approximately 2.2- and 3.0-fold increases in DNA fragmentation were observed, respectively (Table 5). Due to extensive cell necrosis, DNA fragmentation could not be measured at 48 h. Cultured J774A.1 murine macrophage cells were treated with 0, 10, 30 and 50 mg/ml concentrations of chromium(III) picolinate or niacin-bound chromium(III) for 24 h. Approximately 1.0-, 1.2- and 1.7-fold increases in DNA fragmentation were observed following incubation of these cells with 10, 30 and 50 mg/ml concentrations of chromium(III) picolinate, respectively, as compared with the control untreated cells, while under these same conditions approximately 1.1-, 1.2- and 1.3-fold increases in DNA fragmentation were observed following treatment with same concentrations of niacin-bound chromium(III) (Table 6). Concentration-dependent increases in DNA fragmentation were also observed following incubation of cultured J774A.1 macrophage cells with chromium(VI). Following incubation of J774A.1

macrophage cells with 0.20, 0.40, and 0.60 mM concentrations of chromium(VI) for 24 h approximately 1.8-, 2.8- and 2.9-fold increases in DNA fragmentation were observed, respectively, as compared with control cells (Table 10). Genomic DNA fragmentation was determined in the liver and brain tissues obtained from both C57BL/6Ntac and p53 deficient C57BL/6TSG p53 mice. Increases in DNA fragmentation were observed in the liver and brain tissues of both strains of mice at both 0.10 LD50 and 0.50 LD50 doses. Following treatment with acute oral 0.10 LD50 and 0.50 LD50 doses of chromium(VI) approximately 1.4- and 2.2-fold increases in DNA fragmentation were observed in the liver, and 1.3- and 2.1-fold increases in DNA fragmentation in the brain tissues of C57BL/6Ntac mice, respectively, as compared with the control untreated animals (Tables 7 and 8). While under these same conditions, chromium(VI)-induced 3.0- and 4.1-fold increases in DNA fragmentation in the liver, and 2.1- and 4.6-fold increases in DNA fragmentation in the brain tissues of p53 deficient C57BL/6TSG p53 mice, respectively, (Tables 7 and 8). Thus, significantly higher DNA fragmentation was observed in both the liver and brain tissues of p53 deficient C57BL/6TSG p53 mice compared with the corresponding control animals.

3.6. Reduction of MTT Cultured J774A.1 macrophage cells were incubated with 0.20, 0.40 and 0.60 mM concentrations of chromium(VI) for 24 and 48 h, and concentration- and time-dependent effects were studied on the reductions of tetrazolium dye MTT. Thus, the activity of succinate dehydrogenase, a marker of the mitochondrial electron transport chain, was assessed. Following treatment of the macrophage cells with 0.20 and 0.40 mM concentrations of chromium(VI) for 24 h resulted in no significant changes in succinate dehydrogenase activity. While treatment with 0.60 mM concentration of chromium(VI) for 24 h, a small but non-significant increase in succinate dehydrogenase activity, was observed. Incubation of these cells with the three concentrations of chromium(VI) for 48 h resulted

Control C57BL/6Ntac mice and p53-deficient C57BL/6TSG p53 mice were treated with a single oral 0.10 LD50 (19mg/kg body weight) or 0.50 LD50 (95 mg/kg body weight) dose of sodium dichromate [chromium(VI)] and sacrificed after 24 h. Each value is the mean9/S.D. of four mice. Assays were performed, as described in the Section 2. Values with non-identical superscripts are significantly different (P B/0.05).

12.269/1.62c (405) 74.309/10.29b 3.039/0.46a (100) 9.079/0.73b (300) 6.719/0.90a 36.039/5.78 10.069/1.32c (220) 49.869/6.76b 4.579/0.72a (100) 6.499/0.78b (142) 8.319/1.14a 20.369/4.90

20.399/3.16c 3.419/0.27b 1.899/0.26a 2.939/0.38b

Lipid peroxidation (nmoles/mg protein) DNA fragmentation (% control) Superoxide anion (nmoles cytochrome c reduced/15 min/mg/ protein

2.099/0.25a

6.849/0.91c

Chromium(VI) 0.50 LD50 Chromium(VI) 0.10 LD50 Control Chromium(VI) 0.10 LD50 Control

Chromium(VI) 0.50 LD50

p53-deficient (TSG-p53) mice Control (C57BL/6Ntac) mice Assay

Table 7 Hepatic lipid peroxidation, DNA fragmentation, superoxide anion production in sodium dichromate [chromium(VI)] treated control and p53-deficient mice

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

15

in dose-dependent decreases in succinate dehydrogenase activities (Table 10). 3.7. Flow cytometric analyses Following treatment of the human K562 cells with 12.5 mM concentration of chromium(VI) for 24 h, approximately 40% apoptotic cell death was observed (Table 5) in conjunction with a significant decrease in the G2/M phase (data not shown). Due to cell necrosis, no apoptotic cell death was observed in cultured K562 cells following treatment with 25 mM concentration of chromium(VI) for 24 or 48 h, and necrotic cell death was confirmed microscopically (data not shown). 3.8. Chromium(VI)-induced cell morphology and apoptotic cell death in J774A.1 cells Phase contrast microscopy demonstrated that chromium(VI)-induced concentration-dependent changes in cell morphology. With increasing concentration of chromium(VI), cultured J774A.1 cells lost their adhesion to the culture plates and the cells became typically rounded in appearance (data not shown). A concentration-dependent effect was observed. A MEBSTAIN apoptosis kit was used to visualize the effects of 0.20, 0.40 and 0.60 mM concentrations of chromium(VI) on cultured J774A.1 cells, and a concentration-dependent effect was observed (Bagchi et al., 2001). A concentration-dependent increasing index of apoptotic cell death was evident by the increasing intensity and volume of the green fluorescent color in single cells, demonstrating cytoplasmic condensation with intact membrane and organelle structure (data not shown).

4. Discussion This study is focused to determine the comparative effects of hexavalent and trivalent chromium on the enhanced production of oxidative stress as demonstrated by enhanced production of superoxide anion and hydroxyl radicals, increased excretion of urinary lipid metabolites including MDA, FA, ACT and ACON, increased lipid

Control C57BL/6Ntac mice and p53-deficient C57BL/6TSG p53 mice were treated with a single oral 0.10 LD50 (19 mg/kg body weight) or 0.50 LD50 (95 mg/kg body weight) dose of sodium dichromate [chromium(VI)] and sacrificed after 24 h. Each value is the mean9/S.D. of four mice. Assays were performed, as described in the Section 2. Values with non-identical superscripts are significantly different (P B/0.05).

12.559/1.22c (455) 56.809/8.94c 2.769/0.24a (100) 5.819/0.66b (211) 7.349/0.65a 26.949/4.58b 6.209/0.51c (210) 23.839/5.56c

5.889/0.81c

Lipid peroxidation (nmoles/mg protein) DNA fragmentation (% control) Superoxide anion (nmoles cytochrome c reduced/15 min/mg per protein

2.969/0.40a (100) 3.909/0.41b (132) 5.789/0.41a 14.579/2.08b

14.719/1.69c 4.799/0.33b 2.849/0.42b 1.699/0.19a

2.029/0.24a

Control Chromium(VI) 0.10 LD50 Control

Chromium(VI) 0.50 LD50

P53-deficient (TSG-p53) mice Control (C57BL/6Ntac) mice

Chromium(VI) 0.10 LD50

Chromium(VI) 0.50 LD50

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

Assay

Table 8 Brain lipid peroxidation, DNA fragmentation, superoxide anion production in sodium dichromate [chromium(IV)] treated control and p53-deficient mice

16

peroxidation, DNA fragmentation and apoptotic cell death in in vitro and in vivo models. The effects of acute oral doses of 0.10 LD50 and 0.50 LD50 of chromium(VI) were also assessed on increased production of superoxide anion, lipid peroxidation and DNA fragmentation in hepatic and brain tissues of female C57BL/6Ntac and p53deficient C57BL/6TSG p53 mice to evaluate the role of apoptotic regulatory p53 gene. In addition, to compare the relative cytotoxicity of two popular widely used trivalent chromium nutritional supplements, we assessed concentration-dependent effects of chromium(III) picolinate and niacinbound chromium(III) on enhanced production of superoxide anion, lipid peroxidation and DNA fragmentation in cultured J774A.1 murine macrophage cells. Previous studies demonstrated that chromium(VI) induced oxidative stress through enhanced production of ROS and nitric oxide, DNA damage, as well as enhanced excretion of urinary lipid metabolites including MDA, FA, ACT and ACON, activation of protein kinase C in both in vitro and in vivo models (Bagchi et al., 1995a,b, 1997a,b, 1998, 2000, 2001). Results show that hexavalent chromium exhibited more pronounced cytotoxicity at equimolar concentration when compared with trivalent chromium as demonstrated by increased production of ROS by peritoneal macrophages and hepatic mitochondria and microsomes, and enhanced excretion of urinary lipid metabolites (Tables 1 and 4). Urinary excretion of lipid metabolites following administration of low, daily, oral chronic doses of 0.05 LD50 (2.5 mg/kg body weight/day) or 0.20 LD50 (10 mg/kg body weight/day) of sodium dichromate [chromium(VI)] was determined over a period of 90 days. Significant increases in all four lipid metabolites were observed at 0, 30, 60 and 90 days of treatment (Tables 2 and 3). Similar results were also obtained in hepatic mitochondrial and microsomal lipid peroxidation at 0, 30, 60 and 90 days of treatment (Table 9). Hexavalent chromium induced a more drastic effect on p53-deficient C57BL/6TSG p53 mice compared with the control C57BL/6Ntac mice (Tables 7 and 8). The tumor suppressor gene p53 is involved in key responses to genotoxic stress,

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

17

Table 9 Hepatic lipid peroxidation in Sprague
/Dawley rats at 0, 30, 60, and 90 days of treatment with 10 mg sodium dichromate [chromium VI]/kg body weight/day Days of sodium dichromate treatment

Hepatic lipid peroxidation (nmol MDA/mg protein) Mitochondria

0 30 60 90

Microsomes

Control

Treated

Control

Treated

5.129/0.43 6.749/0.66 7.339/0.78 7.879/0.69


/ 10.689/0.88a 17.609/1.03a 16.449/1.42a

3.659/0.24 3.989/0.64 4.509/0.23 5.309/0.64


/ 5.529/0.48a 8.549/0.57a 10.569/0.81a

Female Sprague
/Dawley rats received 10 mg sodium dichromate [chromium(VI)]/kg body weight/day orally in water. Control animals received the vehicle. The animals were sacrificed on days 0, 30, 60, and 90. Each value is the mean9/S.D. of four to six animals. a P B/0.05 with respect to the corresponding control group.

and Schwartz and Rotter (1998) reviewed the known molecular pathways by which the gene controls the cell cycle, with a specific focus on the significance of the p53-mediated checkpoint response for its tumor suppressor function. Posttranslational modification of the p53 protein through phosphorylation is believed to be an important regulatory mechanism of p53 function (Milczarek et al., 1997). Furthermore, the p53 tumor suppressor gene plays a major role in the regulation of the cellular stress response, in part, through the transcriptional activation of genes involved in cell cycle control, DNA repair, and apoptosis (Amundson et al., 1998; Singh et al., 1998). Since p53 is activated in response to DNA damage, and various factors are known to interact to signal and modulate this response, the role of p53 gene was assessed in chromium(VI)-induced toxicity using C57BL/6TSG p53-deficient mice. In the in vitro studies, hexavalent chromium induced oxidative stress in both chronic myelogenous leukemic K562 and J774A.1 murine macrophage cells. A concentration-dependent effect was observed following treatment of the K562 cells with 12.5 and 25 mM concentration of chromium(VI) as demonstrated by enhanced production of superoxide anion and hydroxyl radicals, as well as increased DNA fragmentation and apoptotic cell death (Table 5). Concentrationdependent changes in DNA fragmentation and reduction of tetrazolium dye MTT in cultured murine macrophage cells were observed following

treatment with 0.20, 0.40 and 0.60 mM concentrations of chromium(VI) (Table 10). Concentrationand time-dependent effects of chromium(VI) were assessed on the reduction of tetrazolium dye MTT in cultured J774A.1 cells. MTT assay is a marker of succinate dehydrogenase activity, an index of the mitochondrial electron transport system, which was used to assess cell viability. Only viable cells with intact mitochondria can reduce tetrazolium dye MTT and the amount of MTT reduced is directly proportional to the number of viable cells present. At the 24 h time point, chromium(VI) had no effect on the reduction of MTT, suggesting that little or no loss in cell viability had occurred. However, at the 48 h time point, significant concentration-dependent decreases in MTT reduction were observed. These exposures were accompanied by detachment of dead or dying cells in the macrophage cultures (Table 10). In our previous study, we conducted a comparative study to demonstrate the concentrationdependent effect of chromium(VI) on cultured murine macrophage J774A.1 cells, normal human donor peripheral blood mononuclear cells (HPBM), and human chronic myelogenous leukemic K562 cells (Bagchi et al., 2000). Chromium(VI) induced a cytotoxic effect of macrophage J774A.1 cells at a much lower dose compared with the other two cells. Modulation of intracellular oxidized states was also observed as assessed by laser scanning confocal microscopy.

18

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

Table 10 Concentration-dependent changes in DNA fragmentation and reduction of tetrazolium dye MTT in cultured J774A.1 murine macrophage cells following treatment with sodium dichromate [chromium(VI)] Treatment

DNA fragmentation

Cell viability (%) 24 h

Control Cr(VI) 0.20 mM Cr(VI) 0.40 mM Cr(VI) 0.60 mM

7.259/0.63a (100) 13.239/1.89b (183) 20.159/0.71c (278) 21.049/1.08c (290)

1.079/0.11a 0.999/0.03a 1.149/0.17a 1.279/0.10a

48 h (100) (93) (107) (119)

1.029/0.16a (100) 0.719/0.03b (70) 0.659/0.06b,c (64) 0.589/0.03c (57)

Cultured J774A.1 macrophage cells were treated with either 0.20, 0.40, or 0.60 mM sodium dichromate [chromium(VI)]. See Section 2 for details. All values are reported as mean9/S.D. from four to six replicates. Values with non-identical superscripts are significantly different (P B/0.05).

Chromium(VI) induced a concentration-dependent effect on macrophage J774A.1 cells with maximal effect observed at 0.40 mM concentration. A differential effect of chromium(VI) was observed in HPBM and human K562 cells. No significant modulation of intracellular oxidized states was observed in HPBM cells following treatment with 12.5 mM concentration of chromium(VI) for 24 h, while under these same conditions a 5.7-fold increase in fluorescent intensity was observed in K562 cells. Approximately a 2.2-fold increase in fluorescence intensity was observed following incubation of HPBM cells with 12.5 mM concentration of chromium(VI) for 48 h, while similar conditions induced cell necrosis in cultured K562 cells. Thus, chromium(VI) was shown to induce a more drastic effect on human leukemic cells when compared with human normal cells. Similar results were observed in the production of superoxide anion and hydroxyl radicals. Human K562 cells induced a more dramatic production of superoxide anion and hydroxyl radicals compared with normal HPBM cells. The comparative modulation of intracellular oxidized states and enhanced production of superoxide anion and hydroxyl radicals in these cells, following incubation with chromium(VI), was further correlated with apoptotic cell death. Induction of apoptotic cell death was demonstrated by three sensitive markers, namely, genomic DNA fragmentation, TUNEL assay, and flow cytometry. Apoptosis (programmed cell death) is a widespread phenomenon that plays a crucial role in a myriad of pathophysiological processes, including

neoplastic diseases. The present study demonstrates the toxic effect of chromium(VI) on different cultured cells. Chromium(VI) induced significant DNA fragmentation in cultured J774A.1 cells and the maximal effect was observed at 0.40 mM concentration of chromium(VI). These genomic DNA fragmentation data were further supported by ladder-like fragmentation of genomic DNA and TUNEL assay. At both concentrations and time points, chromium(VI) induced more pronounced DNA fragmentation of human chronic myelogenous leukemic cells compared with normal HPBM cells. Similar results were observed in cell cycle analysis. Chromium(VI) induced no apoptotic cell death on HPBM cells at either concentration or time point, while 40% apoptotic cell death was observed in human K562 cells treated with 12.5 mM concentration of chromium(VI) for 24 h. Necrotic cell death was observed following treatment of the human K562 cells with 12.5 mM concentration of chromium(VI) for 48 h, or incubation with a 25 mM concentration of chromium(VI) for 24 or 48 h (Bagchi et al., 2000, 2001). Trivalent chromium compounds are 1000-fold less toxic than hexavalent chromium compounds, but trivalent chromium compounds may cause toxicity at higher concentrations and/or depending on the ligands attached to it (Barceloux, 1999; Biedermann and Landolph, 1990; Bagchi et al., 1995a, 1997c; Levis et al., 1978; Stohs and Bagchi, 1995). This difference in toxicity is believed to be due to the fact that hexavalent chromium can penetrate biological membranes via non-specific

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

anion carriers more readily than trivalent chromium (Danielsson et al., 1982; Levis et al., 1978). It is important to mention that chromium levels decrease with age and marginal chromium deficiencies appear to be widespread. Only trivalent chromium potentiates the action of insulin both in vitro and in vivo. However, trivalent chromium does not work physiologically as a free element, but rather when it is linked with a suitable ligand such as nicotinic acid (Mertz, 1993). It has been well documented that the form of ligand is the primary influence in chromium’s bioavailability, function and safety. Maximal in vitro activity requires a special chemical form, termed glucose tolerance factor (GTF). The biologically active form of trivalent chromium, GTF, has been identified as a chromium
/nicotinic acid complex and shown to efficiently potentiate insulin function in regulating carbohydrate metabolism and reducing levels of LDL cholesterol (Mertz, 1993). Chromium (only as GTF) is easily absorbed, bioavailable and transported across the placenta, has different tissue distribution and access to biologically important storage depots than other forms of chromium, potentiates insulin better than inorganic chromium salts, and is safe (Mertz, 1993). The Human Risk Assessment Branch of USEPA investigated the safety and toxicity of trivalent chromium. The EPA established a Reference Dose Value (RfD) for Chromium of 1000 mcg/kg per day. This value is 333
/1533 times greater than that of the ESADDI (50
/200 mcg). The RfD is defined as ‘an estimate of a daily exposure to the human population that is likely to be without appreciable risk of deleterious effects during a lifetime.’ Trivalent chromium supplements such as chromium picolinate and niacin-bound chromium(III) have been widely marketed as forms of biologically active chromium that promote healthy blood sugar and blood lipid levels, lean body mass and muscle strength (Anderson et al., 1997; Preuss and Anderson, 1998). In this paper, we have demonstrated the concentration-dependent effects of chromium(III) picolinate and niacin-bound chromium(III) on enhanced production of ROS including superoxide anion and hydroxyl radicals, and DNA fragmentation following incubation

19

with cultured J774A.1 murine macrophage cells. Although both chromium(III) supplements exhibit significantly much less oxidative stress and DNA damage compared with chromium(VI), chromium(III) picolinate exhibited higher production of noxious superoxide anion and increased DNA fragmentation compared with niacin-bound chromium(III). Recently, chromium picolinate has been demonstrated to be mutagenic at the hypoxanthine (guanine) phosphoribosyltransferase locus in Chinese hamster ovary cells at physiological relevant doses, while at the same dose chromium chloride was not mutagenic. Chromium picolinate caused significant cell death, while chromium chloride at the same dose did not cause any cell death or toxicity (Stearns et al., 2002). According to these researchers, coordination of chromium(III) with picolinic acid may make chromium more genotoxic than other forms of chromium(III). In another study, rats were given chronic oral doses of 0, 2.66 (low), 5.32 (medium) or 10 (high) mg/kg per day of chromium picolinate for 21 days. An increase in lipid peroxidation levels in liver and kidney tissues were observed in all chromium picolinate-treated rats. SOD, glutathione peroxidase (GPx) and glutathione (GSH) levels in the hepatic tissues were decreased in all the treated groups, while the hepatic catalase (CAT) level decreased in the high dose group. The kidney SOD and CAT levels also decreased in all treated groups, while GSH and GR levels were reduced in the mid- and high-dose treated groups (Mahboob et al., 2002) In accordance with our findings, a number of recent studies have reported toxic manifestations of chromium picolinate. Chromium picolinate was found to produce chromosome damage 3- to 18fold greater than control levels for soluble doses of 0.050, 0.10, 0.50 and 1.0 mM after 24 h treatment (Stearns et al., 1995a). 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 alone resulted in chromosomal damage and is clastogenic (Beskid et al., 1995; Stearns et al., 1995a,b). Speetjens et al. (1999)

20

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

postulated that chromium picolinate in the presence of reductants and air is capable of generating hydroxyl radicals, which in turn can cleave supercoiled DNA. These results are consistent with other studies, which demonstrate the ability of certain chromium complexes to nick DNA in the presence of peroxide and a reductant or in cells. Chromium picolinate has also been shown to generate hydroxyl radicals from hydrogen peroxide in a pathway independent of added reductant. Due to chromium’s ligand composition and the resulting redox potential, the complex can be reduced readily by abundant biological reductants and generate hydroxyl radicals via Haber
/Weiss and Fenton reactions. The enhanced production of hydroxyl radicals results in appreciable DNA damage (Speetjens et al., 1999). Chromium picolinate was also indicated as the cause of chronic renal failure in a 33-year old woman when she took chromium picolinate 1200
/ 2400 mg/day for 4
/5 months to enhance weight loss. This case was presented with weight loss, anemia, thrombocytopenia, hemolysis, liver dysfunction (aminotransferase enzymes 15
/20 times normal, total bilirubin three times normal), and renal failure (serum creatinine 5.3 mg/dl: blood urea nitrogen (BUN) 152 mg/dl) (Cerulli et al., 1998). The patient had chromium plasma concentrations two to three times normal. All the values returned to a normal level after she discontinued chromium picolinate for 12 months. These authors indicated that chromium picolinate causes serious renal impairment when ingested in excess (Cerulli et al., 1998). In another case study a 49-year old female nurse ingested 600 mg chromium picolinate/ day for 6 weeks for weight reduction (Wasser and Feldman, 1997). Test results had shown normal renal functions 2 years earlier, however, supplementation with chromium picolinate showed a BUN level of 74 mg/dl and serum creatinine level of 5.9 mg/dl. Urinalysis showed a protein level of 30 mg/dl, trace amounts of blood, a leukocyte count of 0
/1 cell per high-power field and an erythrocyte count of 1
/2 cells per high-power field. The 24 h urine protein level was 782 mg. The patient admitted to having ingested chromium picolinate 5 months earlier. From the time of her last normal serum creatinine level until presenta-

tion, the only other medications she received were antihypertensive agents. A renal biopsy showed results consistent with severe chronic active interstitial nephritis. These human case studies suggest that chromium picolinate causes nephrotoxicity with chronic use. A clinical case of acute generalized exanthematous pustulosis was also reported to be caused by chromium picolinate consumption in a 32-year old man (Young et al., 1999), while another clinical case study reports that a 35-year old man developed systemic chronic dermatitis from taking chromium picolinate (Fowler, 2000). Another clinical report demonstrated that a 35 year-old male patient experienced progressively worsening episodes of cognitive, perceptual and motor changes that interfered with his ability to drive a car following ingestion of 200 and 400 mg of chromium picolinate (Huszonek, 1993). Thus, biochemical, physiological, toxicological and behavioral actions of chromium picolinate appeared to be a consequence of the effects of picolinic acid on the central nervous system and other vital target organs (Beskid et al., 1995). In contrast to these findings, a thorough search of the scientific literature reveals superior bioavailability, efficacy and safety for niacin-bound chromium(III) (Crawford et al., 1999; Grant et al., 1997; Lefavi et al., 1993; Olin et al., 1994; Preuss et al., 1995, 1997, 1998, 2000, 2001; Tyson et al., 2000). Taken together, these results in conjunction with our previous studies (Bagchi et al., 1995a,b, 1997a,b, 1998, 2000, 2001; Stohs et al., 2001) demonstrate that hexavalent chromium results in enhanced formation of ROS, including superoxide anion, hydroxyl radicals and nitric oxide, decreased cell viability, increased cellular and genomic hepatic DNA fragmentation, enhanced intracellular oxidized states, membrane damage with leakage of lactate dehydrogenase, activation of protein kinase C, apoptotic and necrotic cell death. These data are consistent with previously published observations regarding the involvement of oxidative stress, apoptotic cell death and modulation of p53 apoptotic regulatory gene in the toxicity and carcinogenicity of chromium(VI). Furthermore, based on extensive scientific, clinical and mechanistic research, and toxicological assess-

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

ment, trivalent chromium compounds are relatively non-toxic and beneficial for human health, although questions concerning the safety of chromium picolinate persist.

Acknowledgements This work was supported in part by grants from the Air Force Office of Scientific Research (#94-10048 and #97-1-0016). The authors thank Dr Walter Kozumbo for scientific discussions and Kristine Strong for technical assistance.

References Amundson, S.A., Myers, T.G., Fornace, A.J., Jr, 1998. Roles of p53 in growth arrest and apoptosis: putting on the brakes after genotoxic stress. Oncogene 17, 3287
/3299. Anderson, R.A., Cheng, N., Bryden, N.A., Polansky, M.M., Cheng, N., Chi, J., Feng, J., 1997. Elevated intakes of supplemental chromium improve glucose and insulin variables in individuals with type 2 diabetes. Diabetes 46, 1786
/ 1791. Babior, B.M., Kipnes, R.S., Curnutte, J.T., 1973. Biological defense mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. J. Clin. Invest. 52, 741
/744. Bagchi, D., Hassoun, E.A., Bagchi, M., Muldoon, D., Stohs, S.J., 1995a. Oxidative stress induced by chronic administration of sodium dichromate [Cr(VI)] to rats. Comp. Biochem. Physiol. 110C, 281
/287. Bagchi, D., Hassoun, E., Bagchi, M., Stohs, S., 1995b. Chromium-induced excretion of urinary lipid metabolites, DNA damage, nitric oxide production, and generation of reactive oxygen species in Sprague
/Dawley rats. Comp. Biochem. Physiol. 110C, 177
/187. Bagchi, D., Bagchi, M., Balmoori, J., Ye, X., Stohs, S.J., 1997a. Comparative induction of oxidative stress in cultured J774A.1 macrophage cells by chromium picolinate and chromium nicotinate. Res. Commun. Mol. Pathol. Pharmacol. 97, 335
/346. Bagchi, D., Bagchi, M., Tang, L., Stohs, S.J., 1997b. Comparative in vitro and in vivo protein kinase C activation by selected pesticides and transition metal salts. Toxicol. Lett. 91, 31
/39. Bagchi, D., Vuchetich, P.J., Bagchi, M., Hassoun, E.A., Tran, M.X., Tang, L., Stohs, S.J., 1997c. Induction of oxidative stress by chronic administration of sodium dichromate [chromium VI] and cadmium chloride [cadmium II] to rats. Free Radic. Biol. Med. 22, 471
/478. Bagchi, D., Tran, M.X., Newton, S., Bagchi, M., Ray, S.D., Kuszynski, C.A., Stohs, S.J., 1998. Chromium and cad-

21

mium induced oxidative stress and apoptosis in cultured J774A.1 macrophage cells. In Vitro Mol. Toxicol. 11, 171
/ 181. Bagchi, D., Joshi, S.S., Bagchi, M., Balmoori, J., Benner, E.J., Kuszynski, C.A., Stohs, S.J., 2000. Cadmium- and chromium-induced oxidative stress, DNA damage, and apoptotic cell death in cultured human chronic myelogenous leukemic K562 cells, promyelocytic leukemic HL-60 cells, and normal human peripheral blood mononuclear cells. J. Biochem. Mol. Toxicol. 14, 33
/41. Bagchi, D., Bagchi, M., Stohs, S.J., 2001. Chromium(VI)induced oxidative stress, apoptotic cell death and modulation of p53 tumor suppressor gene. Mol. Cell. Biochem. 222, 149
/158. Barceloux, D.G., 1999. Chromium. J. Toxicol. Clin. Toxicol. 37, 173
/194. Beskid, M., Jachimowicz, J., Taraszewska, A., Kukulska, D., 1995. Histological and ultrastructural changes in the rat brain following systemic administration of picolinic acid. Exp. Toxicol. Pathol. 47, 25
/30. Biedermann, K.A., Landolph, J.R., 1987. Induction of anchorage independence in human diploid foreskin fibroblasts by carcinogenic metal salts. Cancer Res. 47, 3815
/3823. Biedermann, K.A., Landolph, J.R., 1990. Role of valence state and solubility of chromium compounds on induction of cytotoxicity, mutagenesis, and anchorage independence in diploid human fibroblasts. Cancer Res. 50, 7835
/7842. Cerulli, J., Grabe, D.W., Gauthier, I., Malone, M., McGoldrick, M.D., 1998. Chromium picolinate toxicity. Ann. Pharmacother. 32, 428
/431. Crawford, V., Scheckenbach, R., Preuss, H.G., 1999. Effects of niacin-bound chromium supplementation on body composition in overweight African
/American women. Diabetes Obes. Metab. 1, 331
/337. Danielsson, B.R.G., Hassoun, E., Dencker, L., 1982. Embryo toxicity of chromium: distribution in pregnant mice and effects on embryonic cells in vitro. Arch. Toxicol. 51, 233
/ 245. Fowler, J.F., Jr, 2000. Systemic contact dermatitis caused by oral chromium picolinate. Cutis 65, 116. Gavrieli, Y., Sherman, Y., Ben-Sassen, S.A., 1992. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493
/501. Grant, K.E., Chandler, R.M., Castle, A.L., Ivy, J.L., 1997. Chromium and exercise training: effect on obese women. Med. Sci. Sports Exerc. 29, 992
/998. Huszonek, J., 1993. Over-the-counter chromium picolinate. Am. J. Psychiatry 150, 1560
/1561. Jones, P., Kortenkamp, A., O’Brien, P., Wang, G., Yang, G., 1991. Evidence for the generation of hydroxyl radicals from a chromium(V) intermediate isolated from the reaction of chromate and glutathione. Arch. Biochem. Biophys. 286, 652
/655. Kawanishi, S., Inoue, S., Sano, S., 1986. Mechanism of DNA cleavage induced by sodium chromate(VI) in the presence of hydrogen peroxide. J. Biol. Chem. 261, 5952
/5958.

22

D. Bagchi et al. / Toxicology 180 (2002) 5
/22

Kawanishi, S., Hiraku, Y., Murata, M., Oikawa, S., 2002. The role of metals in site-specific DNA damage with reference to carcinogenesis. Free Radic. Biol. Med. 32, 822
/832. Lefavi, R.G., Wilson, G.D., Keith, R.E., Anderson, R.A., Blessing, D.L., Hames, C.G., McMillan, J.L., 1993. Lipidlowering effect of a dietary chromium(III) nicotinic acid complex in male athletes. Nutr. Res. 13, 239
/249. Levis, A.G., Bianchi, V., Tamino, G., Pegoraro, B., 1978. Cytotoxic effects of hexavalent and trivalent chromium on mammalian cells in vitro. Br. J. Cancer 37, 386
/396. Mertz, W., 1993. Chromium in human nutrition: a review. J. Nutr. 123, 626
/633. Milczarek, G.J., Martinez, J., Bowden, G.T., 1997. p53 Phosphorylation: biochemical and functional consequences. Life Sci. 60, 1
/11. Mahboob, M., McNeil, L., Tolliver, T., Ogden, L., 2002. Effects of chromium picolinate on antioxidant enzyme levels in rats. Toxicol. Sci. 66 (1
/S), 32 (Abstract 154). Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J. Immunol. Methods 65, 55
/63. Norseth, T., 1981. The carcinogenicity of chromium. Environ. Health Perspect. 40, 121
/130. Olin, K.L., Stearns, D.M., Armstrong, W.H., Keen, C.L., 1994. Comparative retention/absorption of 51chromium (51Cr) from 51Cr chloride, 51Cr nicotinate and 51Cr picolinate in a rat model. Trace Elem. Electrocytes 11, 182
/186. Patierno, S.R., Banh, D., Landolph, J.R., 1988. Transformation of C3H/10T1/2 mouse embryo cells to focus formation and anchorage independence by insoluble lead chromate but not soluble calcium chromate: relationship to mutagenesis and internalization of lead chromate particles. Cancer Res. 48, 5280
/5288. Preuss, H.G., Anderson, R.A., 1998. Chromium update: examining recent literature 1997
/1998. Curr. Opin. Clin. Nutr. Metab. Care 1, 509
/512. Preuss, H.G., Gondal, J.A., Bustos, E., Bushehri, N., Lieberman, S., Bryden, N.A., Polansky, M.M., Anderson, R.A., 1995. Effects of chromium and guar on sugar-induced hypertension in rats. Clin. Nephrol. 44, 170
/177. Preuss, H.G., Grojec, P.L., Lieberman, S., Anderson, R.A., 1997. Effects of different chromium compounds on blood pressure and lipid peroxidation in spontaneously hypertensive rats. Clin. Nephrol. 47, 325
/330. Preuss, H.G., Jarrell, S.T., Scheckenbach, R., Lieberman, S., Anderson, R.A., 1998. Comparative effects of chromium, vanadium and gymnema sylvestre on sugar-induced blood pressure elevation in SHR. J. Am. Coll. Nutr. 17, 116
/123. Preuss, H.G., Wallerstedt, D., Talpur, N., Tutuncuoglu, S.O., Echard, B., Myers, A., Bui, M., Bagchi, D., 2000. Effects of niacin-bound chromium and grape seed proanthocyanidin extract on the lipid profile of hypercholesterolemic subjects: a pilot study. J. Med. 31, 227
/246. Preuss, H.G., Montamarry, S., Echard, B., Scheckenbach, R., Bagchi, D., 2001. Long-term effects of chromium, grape seed extract, and zinc on various metabolic parameters of rats. Mol. Cell. Biochem. 223, 95
/102.

Prodczasy, J.J., Wei, R., 1988. Reduction of iodonitrotetrazolium violet by superoxide radicals. Biochem. Biophys. Res. Commun. 150, 1294
/1301. Schwartz, D., Rotter, V., 1998. p53-Dependent cell cycle control: response to genotoxic stress. Cancer Biol. 8, 326
/ 336. Shi, X., Dalal, N.S., 1989. Chromium(V) and hydroxyl radical formation during the glutathione reductase-catalyzed reduction of chromium(VI). Biochem. Biophys. Res. Commun. 163, 627
/634. Shi, X., Dalal, N.S., 1990. On the hydroxyl radical formation in the reaction between hydrogen peroxide and biologically generated chromium(V) species. Arch. Biochem. Biophys. 277, 342
/350. Singh, J., Carlisle, D.L., Pritchard, D.E., Patierno, S.R., 1998. Chromium-induced genotoxicity and apoptosis: relationship to chromium carcinogenesis. Oncol. Rep. 5, 1307
/1318 (Review). Speetjens, J.K., Collins, R.A., Vincent, J.B., Woski, S.A., 1999. The nutritional supplement chromium(III) tris(picolinate) cleaves DNA. Chem. Res. Toxicol. 12, 483
/487. Stearns, D.M., Belbruno, J.J., Wetterhahn, K.E., 1995a. A prediction of chromium(III) accumulation in humans from chromium dietary supplements. FASEB J. 9, 1650
/ 1657. Stearns, D.M., Wise, J.P., Sr, Patierno, S.R., Wetterhahn, K.E., 1995b. Chromium(III) picolinate produces chromosome damage in Chinese hamster ovary cells. FASEB J. 9, 1643
/1648. Stearns, D.M., Silveira, S.M., Wolf, K.K., Luke, A.M., 2002. Chromium(III) tris(picolinate) is mutagenic at the hypoxanthine (guanine) phosphoribosyltransferase locus in Chinese hamster ovary cells. Mut. Res. 513, 135
/142. Stohs, S.J., Bagchi, D., 1995. Oxidative mechanisms in the toxicity of metal ions. Free Radic. Biol. Med. 18, 321
/336. Stohs, S.J., Bagchi, D., Hassoun, E.A., Bagchi, M., 2001. Oxidative mechanisms in the toxicity of chromium and cadmium ions. J. Environ. Pathol. Toxicol. Oncol. 20, 77
/ 88. Telford, W., King, L., Fraker, P., 1991. Evaluation of glucocorticoid-induced DNA fragmentation in mouse thymocyte by flow cytometry. Cell Prolif. 24, 447
/459. Tyson, D.A., Talpur, N.A., Echard, B.W., Bagchi, D., Preuss, H.G., 2000. Acute effects of grape seed extract and niacinbound chromium on cardiovascular parameters of normotensive and hypertensive rats. Res. Commun. Pharmacol. Toxicol. 5, 91
/106. von Burg, R., Liu, D., 1993. Chromium and hexavalent chromium. J. Appl. Toxicol. 13, 225
/230. Wasser, W.G., Feldman, N.S., 1997. Chronic renal failure after ingestion of over-the-counter chromium picolinate. Ann. Intern. Med. 126, 410. Young, P.C., Turiansky, G.W., Bonner, M.W., Benson, P.M., 1999. Acute generalized exanthematous pustulosis induced by chromium picolinate. J. Am. Acad. Dermatol. 41, 820
/ 823.