Reactive oxygen species produced from chromate pigments and ascorbate.

The reactions of various chromate pigments and ascorbate were investigated by an ESR spin trapping technique. Production of Cr(V) was detected directly and productions of very electrophilic reactive oxygen species (ROS) was detected via the oxidation of formate. We demonstrated previously that both dissolved oxygen and Cr (V) were essential in the production of ROS in this system, and that ROS production was inhibited by catalase. We studied here the effect of solubility of different chromate pigments: sodium, calcium, strontium, basic zinc, basic lead supported on silica, and lead and barium chromates on the production of ROS in buffered medium and cell culture medium (Dublecco's Modified Eagle medium + fetal calf serum). Sodium, calcium, basic zinc, and basic lead chromates were active in the production of ROS in presence of cell culture medium, whereas lead and barium chromates were inactive.


Introduction
In a recent article (1) we demonstrated that the reaction of soluble Cr(VI) with ascorbate in aqueous aerated medium produces a very electrophilic species of oxygen capable of oxidizing formate to carboxylate radicals. While Cr(V) was essential to observe the oxidative behavior, hydroxyl radicals were not detected in the medium. We suggested that the reactive oxygen species could be a Cr(V)-superoxo complex. These previous findings support classical theories of carcinogenicity implying strongly electrophilic species as DNA-damaging agents (2), and give further evidence for the role of oxidative processes in chromium carcinogenesis (3). The purposes of the present article are to check if the previously proposed mechanism with soluble chromate can work with less soluble or relatively insoluble chromate pigments used in industry; and, in an attempt to validate this mechanism in the biological environment, to study the effect of a typical cell-culture medium on the kinetics of Cr(V) and reactive oxygen species (ROS) formation. We also report

Results and Discussions
The production of ROS and Cr(V) from chromate pigments and ascorbate is presented in Table 1. To probe the reactivity of these solid chromate pigments, we chose to work at low concentration in ascorbate (1 mM), because an excess of this reagent would simply lead to the disappearance of any ROS eventually formed. We also studied the effect of the DMEM/FCS culture medium, which is known to increase the solubility of chromate pigments (4 In general DMEM/FCS is a weakly complexing and reducing medium, and its effect is not very apparent at low Cr(VI) and ascorbate concentrations. Results with 10 mM soluble sodium chromate and 10 mM ascorbate (instead of 1 mM) reveal somewhat lower ROS but higher Cr(V) production in the presence of DMEM/FCS; however, almost all moderately to poorly soluble chromate pigments show no ROS production in the presence of 10 mM ascorbate (data not shown).
It seems that basic zinc chromate yields significantly more Cr(V) than any other chromate ( Table 1), irrespective of the presence of DMEM/FCS. It is possible that the Zn(II) ion thus has a stabilizing effect on Cr(V). This may help explain the synergistic effect between Cr(VI) and Zn(II) observed for SHE cell transformation (6).
Kinetic results of the reaction between soluble chromate with ascorbate, in pres-  Figure 1. The halftimes reported in ordinate are apparent, because they represent an equilibrium between formation and disappearance of the radical species. At 0% DMEM/FCS, we observe a T1/2 of 50 min for the spin adduct [DMPO-COO-], which agrees reasonably well with the value of 60 min obtained by Zalma (7). We observed for Cr(V) a T112 of 20 min, indicating that it may form a complex with the ascorbate radical produced during this reaction (1,8).
While increasing the DMEM/FCS concentration, we observed a gradual augmentation of the T112 values, which became much more pronounced as the concentration of 50% was reached. T112 for Cr(V) increased more steadily than for [DMPO-COO-] , indicating complexation of Cr(V) by DMEM/FCS. The kinetics of ROS and Cr(V) production are sufficiently slow to permit their study in biological systems. At 100% DMEM/FCS, we can extrapolate that both T1/2 values will reach many hours.
The reaction of chromate with ascorbate was chosen because it reacts rapidly and gives a good yield of ROS, and also because ascorbate is an important reductant of chromate in biologic systems (9). However, other biologic reductants can lead to ROS in the presence of chromate in aerated medium. Table 2 suggests that ROS are also produced by a reaction of Cr(VI) with glutathione, y-L-glutamyl-L-cysteinylglycine (GSH) and j-nicotinamide adenine dinucleotide phosphate (NADPH). Both reactions lead to Cr(V) long-lived complexes (10,11). However, the reaction of cysteine with chromate, which produces only short-lived Cr(V) (12), does not lead to formate oxidation in our experimental conditions. These results suggest a correla-tion between the production of ROS in these reactions and the ability to stabilize Cr(V), such stabilization being provided whether by the reductants or their conjugated oxidized form. It seems reasonable to envisage an interaction between paramagnetic 02 and Cr(V), and then appearance of ROS. For ascorbate, we can suggest the following reactions:  [3] where HA-= ascorbate, A = ascorbate radical, A = dehydroascorbate.
We suggest the formation of a Cr(V)...(O02) complex in the Cr(VI)ascorbate reaction because we did not observe -OH in this case (1). For GSH and NADPH, the mechanism is still to be elucidated and may be different than the ascorbate mechanism. -OH formation is very possible with GSH and NADPH, and may involve the reaction of H202 with Cr(V) in a Fentonlike reaction (13).

Conclusions
The reaction of ascorbate with various chromate pigments produces ROS as evidenced by formate oxidation in aqueous solution at 37°C. The ROS production seems closely related to the solubility of the pigments. The presence of a cell-culture medium (DMEM/FCS) has a measurable effect in terms of kinetics of Cr(V) and carboxylate radical production. This is important because the observed half-lives in presence of DMEM/FCS are of sufficient magnitude to allow biologic manifestations (eventually cancer) to occur. The reduction of Cr(VI) leading to ROS requires that some degree of stabiliza-Environmental Health Perspectives tion of the produced Cr(V) occurs. The reduction by cysteine, which does not provide such stabilization, logically does not produce ROS capable of oxidizing formate. Another condition for ROS production would be that the Cr(V) ligands possess some degree of lability, but it is generally recognized that both Cr(VI) and Cr(V) complexes are subject to ligand exchange reactions (14,15), thus permitting the redox couple Cr(VI)/Cr(V) to act as a catalyst.
This article supports mechanistic considerations relative to the appearance of cancer from chromate exposure. It rests on classic theories of carcinogenicity that imply strongly that electrophilic species (including ROS) cause primary DNA damage. The origin of ROS is molecular oxygen, not hydrogen peroxide. We con-sider dissolved oxygen activation mechanisms very important for cancer-causing oxidative damage because oxygen is present in every cells of the body, whereas H202 is produced only in very few cells like macrophages. For chromates, ROS production originating from dissolved oxygen may arise from a variety of biologic reductants, including ascorbate, glutathione, and NADPH.