Interaction between cadmium and selenium in rat plasma.

The metabolism of (75)Se-labeled SeO(3) (2-) and its conversion by intact rat erythrocytes in vitro to a form which complexes with Cd and plasma proteins were studied. By utilizing both excess SeO(3) (2-) and N-ethylmaleimide to lower erythrocyte reduced glutathione (GSH) concentrations, it was shown that the uptake and metabolism of SeO(3) (2-) were GSH-dependent, the probable intermediate being glutathione selenotrisulfide (GSSeSG). Secondary release of selenium by rat erythrocytes had no relation to the erythrocyte transport of oxidized glutathione (GSSG). While fluoride depressed and chromate increased GSSG transport, chromate, a glutathione reductase inhibitor, decreased selenium release. This release appeared to be secondary to a reaction catalyzed by glutathione reductase. The similarity of I(50) values for chromate's inhibition of glutathione reductase and for the inhibition of selenium release further suggested a relationship between these two events. H(2)Se or a similar product of GSSeSG reduction is proposed to be the active product of SeO(3) (2-) metabolism by rat erythrocytes. By use of gel-filtration and ion-exchange methods it was noted that the incubation of H(2)Se with cadmium and plasma produced a Cd-Se complex indistinguishable from that produced by incubation of Cd, SeO(3) (2-), plasma, and erythrocytes in vitro, or that noted following the administration of Cd and SeO(3)in vivo. A mechanism whereby the tissue distribution and toxicity of cadmium are altered by selenium is suggested.


Introduction
Following the observation of Kar, Das, and Mukeri (1) that the testicular lesions produced by cadmium could be prevented by the administration of selenium compounds, a number of workers have demonstrated the protective action of selenium compounds, especially of selenite (Seo32-), on many of the biological effects of cadmium. The reversal of selenium toxicity by cadmium has also been demonstrated (2).
Although selenium completely protected against cadmium-induced testicular damage in rats, the uptake of cadmium by testicular tissue was increased (3). This suggested that selenium diverted cadmium from its usual target in the testes. Cadmium concentrations in other tissues were also altered, but the most pronounced change was in plasma, which showed a 30to 50-fold increase in cadmium levels over control (4,5). The distribution of cadmium in the soluble proteins of tissues was *  also altered by selenium (6,7), and a Cd-Secontaining moiety appeared which moved at the void volume of G-75. The Cd-Se-containing fractions of plasma were of higher molecular weight. Some preliminary data on the interaction of cadmium and selenium have been presented previously by the present authors (8). This paper provides more details on these interactions.
Cadmium and selenium given to rats simultaneously increased the plasma levels of both over the levels seen when each was given alone. The peak level is seen to be attained at 4 hr. The cadmium and selenium are bound to plasma proteins with an atomic ratio of approximately 1. Initially two peaks of 130,000 daltons and 330,000 daltons are seen. The lower molecular weight peak is at a maximum at 1 hr and declines thereafter with the concomitant appearance of a 420,000 dalton peak. The latter peak is at a maximum at 12 hr. The 300,000 dalton peak appears to persist for the 24-hr time period studied.

Results and Discussion
The Cd-Se complex did not form when Cd and SeO32were incubated with plasma in vitro. Upon August 1978 incubation of cadmium and selenium with plasma and washed red cells, the plasma profile showed a single peak containing equimolar amounts of cadmium and selenium and a molecular weight of 130,000. When SeO32was incubated with plasma and red cells, selenium was very rapidly taken up by the red cells and then released over the course of a few minutes into the plasma, presumably in the form of a metabolite. The time course of this process for a number of starting concentrations of selenium is shown in Figure 1. While the release is almost complete at low concentrations of SeO32-(0.5 nmole/ml erythrocytes), this is not so at high concentrations (0.5 ,mole/ml erythrocytes). One limiting factor for the release was the availability of binding sites in the plasma. Very little selenium was released when 2.5 x 10-7 mole of SeO32was incubated with 1 ml of red cells in phosphate-saline buffer, pH 7.4. The selenium released increased from 30%o at equal volumes of red cells and plasma to 90o at a 7-fold excess of plasma over red cells.
The uptake, subsequent metabolism, and release of selenium by rat erythrocytes were dependent on the presence of reduced glutathione (Fig. 2). As the concentration of SeO2-in the preparation was increased, there was a depletion of erythrocyte GSH, and the amount of Se released into the plasma decreased in parallel. However, at higher SeO3  that metabolism had not taken place. Similar results were obtained when N-ethylmaleimide was used to deplete erythrocyte GSH. It seems likely that SeO 2-follows the reductive metabolism, as proposed by Ganther (9) and shown in Eq. (1) H2SeO3 + 4SGH -. GSSeSG + GSSG + 3H20 (1) Sandholm (10) postulated that selenium was released from erythrocytes as a glutathione complex. Jenkins and Hidiroglou (11) suggested that most of the selenium released from bovine erythrocytes in vitro was in the form of GSSeSG. It was proposed that the mechanism of release was similar to the transport of GSSG.
Our conclusion is that the release of selenium is not related to the release of GSSG and that the form of selenium released is not GSSeSG.
Fluoride ion, an inhibitor of GSSG release, had no effect on selenium release from erythrocytes. However, chromate, an inhibitor of glutathione reductase (12,13), while increasing the GSSG levels of both plasma and red cells, can bring about complete inhibition of selenium release from red cells (Fig. 3).
Within erythrocytes GSSG is normally maintained at very low levels by reduction by glutathione reductase and NADPH (14). Release of GSSG from erythrocytes could only be demonstrated when glucose or another source of reducing equivalents was withdrawn or glutathione reductase was inhibited by chromate (13). In cell-free preparations GSSeSG has been found to be a good substrate for yeast glutathione reductase (15)  to the reduction of selenium to the -2 oxidation state as H2Se or a similar selenide compound. It seems likely, therefore, that under physiological conditions GSSeSG would be further reduced by glutathione reductase. The similarity between the 150 values for chromate inhibition of glutathione reductase and inhibition of selenium release (0.35 and 0.25 ,umole of chromate/pml of erythrocytes, respectively) may also suggest that selenium metabolism is linked to glutathione reductase. As noted above, SeO2did not produce a complex with cadmium when incubated with plasma in the absence of erythrocytes. Neither did GSSeSG incubated with plasma. However, bubbling H2Se [generated by the method of Diplock et al. (16)] produced a single Cd-Se peak with an atomic ratio 1 and a molecular weight 130,000. This peak is indistinguishable from that produced by incubating Cd, Se02-, plasma, and erythrocytes. It is also indistinguishable from the lower molecular weight peak seen in rat plasma following administration of Cd and SeO3 in vivo, which is the predominant form of the Cd-Se complex shortly after administration.
If these three peaks were applied separately to QAE A-50 ion-exchange columns, each was eluted at the same portion of the gradient (Fig. 4). The Cd-Se complex was not associated with a major protein peak and had a charge density slightly less negative than albumin. The complex was apparently stable under the conditions of chromatography, which was carried out at pH 6.5.
When the G-150 gel chromatography was carried out at pH 6.5 instead of at pH 8 as previously, the Cd-Se complex formed by bubbling H2Se into plasma containing cadmium or by incubating SeO3 with cadmium, plasma, and erythrocytes showed a molecular weight of 330,000. A similar pattern of Cd-Se distribution was found following the chromatography of plasma from rats injected with Cd and SeO2-1, 12, or 24 hr before sacrifice. These results were taken as evidence that the Cd-Se complex formed under these conditions in vitro was the same as formed in vivo. The metabolic path from SeO2to H2Se has been documented by Ganther (15) and Hsieh and Ganther (17). This mechanism seems to exist both in vitro and in vivo, and seems to be the one whereby the tissue distribution of cadmium is altered and its toxicity reduced.
Only the 130,000 dalton form is produced by SeO2-in the presence of erythrocytes and by H2Se in their absence. The aggregation may be demonstrated by reducing the pH from 8.0 to 6.5 in gel chromatography.
Whether this or a similar complex is formed in compartments other than plasma is not known. If it is, then a large proportion of the cadmium dose may be sequestered in this manner. At 4-5 hr after simultaneous selenium and cadmium administration as much as 20%o of the dose is in the form of the complex in plasma alone. Because of the rapid metabolism of selenium in vivo and by erythrocytes in vitro the complex will only be formed if the selenium is not given too far ahead of the cadmium. The efficacy of SeO2-in protecting against cadmium toxicity is much greater than that of SeO2-and the seleno amino acids. Presumably these compounds must first be converted to SeO2-or traverse an alternate pathway to H2Se.
The fate of the Cd-Se complex has not been studied. The amount of the complex is about half its peak value at 24 hr. The excretion of cadmium is not significantly increased (18) and cadmium inhibits the urinary excretion of selenium, although six times as much selenium as cadmium appears in the excreta at 48 hr. The cadmium may deposit slowly in other tissues as the complex is decomposed. The liver burden of cadmium is less when selenium is given simultaneously so that other tissues must have higher levels. The slow rate of release of cadmium may allow secondary protection such as is afforded by the synthesis of metallothionein.