Ceruloplasmin alters the tissue disposition and neurotoxicity of manganese, but not its loading onto transferrin

Abstract

Manganese (Mn) is a redox-active element, and whereas its uptake, disposition, and toxicity in mammals may depend in part on its oxidation state, the proteins affecting manganese oxidation state and speciation in vivo are not well known. Studies have suggested that the oxidase protein ceruloplasmin (Cp) mediates iron and manganese oxidation and loading onto plasma transferrin (Tf), as well as cellular iron efflux. We hypothesized that ceruloplasmin may also affect the tissue distribution and eventual neurotoxicity of manganese. To test this, aceruloplasminemic versus wild-type mice were treated with a single i.p. (54)Mn tracer dose, or elevated levels of manganese subchronically (0, 7.5, or 15 mg Mn/kg s.c., three doses per week for 4 weeks), and evaluated for transferrin-bound manganese, blood manganese partitioning, tissue manganese disposition, and levels of brain glutathione, thiobarbituric acid reactive substances (TBARS), and protein carbonyls as measures of oxidative stress, and open arena activity. Results show that ceruloplasmin does not play a role in the loading of manganese onto plasma transferrin in vivo, or in the partitioning of manganese between the plasma and cellular fractions of whole blood. Ceruloplasmin did, however, affect the retention of manganese in blood and its distribution to tissues, most notably kidney and to a lesser extent brain and lung. Results also indicate that ceruloplasmin interacted with chronic elevated manganese exposures to produce greater levels of brain oxidative stress. These results provide evidence that metal oxidase proteins play an important role in altering neurotoxicity arising from elevated manganese exposures.

FIG. 1.

Seventy percent of the collected ⁵⁴Mn activity was associated with transferrin in HPLC fractions 3 and 4 (transferrin identified by comparison with transferrin standards analyzed by HPLC and SDS-PAGE). Shown is a representative anion-exchange HPLC chromatogram of dialyzed Cp^+/+ mouse plasma. Collected fractions (#1–10) indicated by vertical hash marks on x-axis. ⁵⁴Mn activity within collected fractions (total cpm, right y-axis) indicated as bars above the fraction number. HPLC used a three-step linear salt gradient (30 ml of 0–0.08M NaCl, 30 ml of 0.08–0.10M NaCl, 30 ml of 0.10–0.50M NaCl in 20mM tris, pH 8.65) at a flow rate of 1 ml/min, and absorbance was monitored at 280 nm. Below each fraction is the corresponding analyses with SDS-PAGE with Coomassie stain. Molecular weight marker (M), purified human transferrin (Tf), and undialyzed plasma (P) are also shown.

FIG. 2.

There is no difference (F_1,13 = 0.20, p = 0.67) in transferrin-bound manganese between Cp^+/+ and Cp^−/− mice (mean ± SE, n = 7 per group). Data reflect cpm of transferrin-containing HPLC fractions (fractions 3 and 4 in Fig. 1) as a percent of cpm in all HPLC fractions collected.

FIG. 3.

(a) Cp^−/− mice (hatched bars) have more ⁵⁴Mn than Cp^+/+ mice (solid bars) in whole blood, cellular fraction, and plasma (mean ± SE, n = 6–7, *p < 0.05, **p < 0.01). Blood was collected and separated into plasma and cellular fractions 1 h after i.p. injection of 250 μCi ⁵⁴Mn/kg bw. (b) There is no ceruloplasmin genotype difference in ⁵⁴Mn partitioning between plasma and cellular fractions (mean ± SE, n = 6–8, p = 0.96) (% of Mn in the cellular fraction was calculated as hematocrit × specific activity of cellular fraction/specific activity of whole blood).

FIG. 4.

⁵⁴Mn in tissues of Cp^+/+ mice (solid bars) and Cp^−/− mice (hatched bars) (mean ± SE, n = 7, *p < 0.05, #p < 0.10). Cp^−/− mice contained significantly more ⁵⁴Mn in kidneys than Cp^+/+, and Cp^−/− mice trended toward more ⁵⁴Mn in lungs and brain than Cp^+/+ mice. Tissues were collected 1 h after i.p. injection of 250 μCi ⁵⁴Mn/kg bw.

FIG. 5.

Tissue manganese concentrations after 4 weeks of manganese treatment in Cp^+/+ (solid bars) and Cp^−/− mice (hatched bars); males (left panels) and females (right panels) are shown separately. (a) Blood manganese levels (μg Mn/kg wet weight) increase with manganese treatment (p < 0.001), but there are no significant differences between ceruloplasmin genotypes (p = 0.81) or genders (p = 0.23). (b) Brain manganese levels (mg Mn/kg dry weight) increase with treatment (p < 0.001), and are higher in male mice (p < 0.05), but there is no difference between ceruloplasmin genotypes (p = 0.17). (c) Liver manganese levels (mg Mn/kg dry weight) increase with treatment (p < 0.001), and are higher in females than males (p < 0.001); there is not a significant ceruloplasmin genotype effect (p = 0.60). All bars are mean ± SE (n = 3–4).

FIG. 6.

Lipid oxidation (a) and total (oxidized + reduced) glutathione levels (b) in cerebellum of male Cp^+/+ (solid bars) and Cp^−/− mice (hatched bars) after 4 weeks of manganese treatment (cumulative manganese doses shown on x-axis; mean ± SE, n = 3–4). (a) Two-way ANOVA analysis shows significant manganese treatment (p < 0.05) and ceruloplasmin genotype effects in lipid oxidation (p < 0.05), with a significant difference between genotypes for the 180 mg Mn/kg dose, and a significant difference between the 0 and 90 mg Mn/kg treatment groups for Cp^+/+ mice (*p < 0.05, Fisher LSD post hoc test applied to Mn treatments within genotypes and to genotypes within Mn treatments). (b) Two-way ANOVA analysis shows a significant overall genotype effect (p < 0.05), but not an overall manganese treatment effect (p = 0.92). There was a significant difference between genotypes within the 90 mg Mn/kg treatment group (p < 0.05, Fisher LSD post hoc test).

FIG. 7.

Open arena activity of male Cp^+/+ (solid bars) and Cp^−/− mice (hatched bars) after 4 weeks of manganese treatment (mean ± SE, n = 3–4). Activity was measured 6 days after the final manganese treatment, and is reported as distance traveled over 35 min (from 5 to 40 min). There was a significant manganese treatment effect (p < 0.001), but no genotype effect (p = 0.60), based on two-way ANOVA.