The selenoprotein P/ApoER2 axis facilitates selenium accumulation in selenoprotein P-accepting cells and confers prolonged resistance to ferroptosis

Abstract

The essential trace element selenium (Se) plays a significant role in redox homeostasis, while Se is very reactive and has a potent toxicity. Understanding the molecular machinery that supports Se metabolism is important for the both physiological and pathophysiological context. Incorporated Se is translated/transformed in the liver into selenoprotein P (SeP; encoded by Selenop), an extracellular Se carrier protein that effectively transports Se to the cells via the binding to its receptor apolipoprotein E receptor 2 (ApoER2), which is taken up by cells. The present study shows that SeP is a source of Se that accumulates intracellularly and can be utilized for prolonged periods under Se-deficient conditions. In cultured cells (RD and SH-SY5Y), glutathione peroxidase (GPX) expression induced by Se supply via the SeP/ApoER2 pathway was maintained longer during Se deficiency than inorganic Se, which was promoted by ApoER2 overexpression. SeP-deficient mice showed a faster decline in brain Se levels when fed a Se-deficient diet. Preserved GPX expression induced by this SeP/ApoER2 axis contributed to oxidative stress and ferroptosis resistance, suggesting that this redundant Se metabolism contributes to prolonged Se utilization and cytoprotection.

Keywords: Apolipoprotein E receptor 2; Ferroptosis; Glutathione peroxidase; Selenium; Selenoprotein P.

Fig. 1

Assessment of Se metabolism through GPX induction using selenite or SeP as Se sources in RD cells. (A) RD cells were treated with equimolar of Se sources (selenite 100 nM or SeP 0.5 μg/mL) for 24 h. The cells were washed with PBS and collected by 0.1 % Trypsin/EDTA. After the wash of the pellets by PBS, the cells were ashed with 70 % nitric acid, and the Se content was determined by ICP-MS. Mean + S.D., n = 3, ns; not significant, Tukey's t-test. (B) RD cells were treated with selenite 100 nM and SeP 0.5 μg/mL to the indicated treatment times and WB was performed. (C, D) Quantitative values of the (C) GPX1 and (D) GPX4 bands of (B) were determined from triplicate repeated data independently, and corrected for GAPDH, respectively, and shown relative to the 24 h point of SeP 0.5 μg/mL as 1. Mean + S.D., n = 3, vs control, ∗P < 0.05, ∗∗P < 0.01, Dunnett's test. (E) ApoER2 expressing plasmid (ApoER2 OE) or empty vector was transfected to RD cells for 24 h. After that the cells were treated with selenite 100 nM and SeP 0.5 μg/mL for 24 h and the Se content was determined by ICP-MS. Mean + S.D., n = 3, ns; ∗∗P < 0.01, Tukey's t-test. loq indicates below limit of quantification. (F) ApoER2 transfected cells were treated with the indicated concentrations of selenite and SeP and WB was performed. The band of ApoER2 and SeP were too high to evaluate accurately, thus low exposure and high exposure were shown. (G, H, I) Quantitative values of the (G) SeP, (H) GPX1 and (I) GPX4 bands were determined from triplicate independent data, corrected for GAPDH, respectively, and shown relative to the point of ApoER2 OE, SeP 5 μg/mL as 1. The band intensity of SeP were acquired from low exposure images. Mean + S.D., n = 3, vs control, ∗P < 0.05, ∗∗P < 0.01, Dunnett's test.

Fig. 2

Subcellular localization of SeP in RD cells. (A) RD cells were seeded on the cover glass and grown for 24 h. SeP (0.5 μg/mL) was treated for 24 h and stained with EEA1 (red), SeP (green) and Hoechst (blue). (B) The cells were stained with RAB7 (red), SeP (green) and Hoechst (blue). (C) The cells were stained with RAB11 (red), SeP (green) and Hoechst (blue). (D) The cells were stained with LAM2 (red), SeP (green) and Hoechst (blue). The scale bar indicates 5 μm. (E) RD cells or ApoER2 over-expressing RD cells (F) were treated with SeP 0.5 μg/mL for 24 h and cell lysates were separated using density gradient centrifugation. Each fractions were subjected to WB.

Fig. 3

Effect of Se-deficiency on Se-metabolism in SeP knockout mice. (A) Experimental design of animal experiment with Se-deficiency. (B) WT mice grown on normal diet (CE2) were fed a Se-deficient diet for 1–3 weeks and plasma SeP expression levels were checked by WB. (C) WB of serum; quantitative values of SeP were determined and corrected for CBB, respectively. Mean ± S.D., n = 3–4, vs CE2, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, Dunnett's test. (D) The cerebrum cortex was ashed with 70 % nitric acid, and the Se content was determined by ICP-MS. Mean ± S.D., n = 3–4, ∗P < 0.05, Dunnett's test. (E) WT and SeP KO mice grown on normal diet (CE2) were fed a Se-deficient diet for 2 weeks and collected the blood and serum was obtained. WB for mouse selenoprotein P (mSeP) were performed. (F) Band quantification values of mSeP were determined and corrected for Coomassie Brilliant Blue (CBB) staining, respectively. Mean ± S.D., n = 4–5, ∗∗P < 0.01, Welch's t-test. (G) The cerebrum cortex or WT and SeP KO mice were ashed with 70 % nitric acid, and the Se content was determined by ICP-MS. Mean ± S.D., n = 4–5, ∗P < 0.05, ∗∗P < 0.01, Tukey'test. (H) Data from D quantified the rate of decrease in Se content due to 2 weeks of Se deficiency. Mean ± S.D., n = 4–5, ∗P < 0.05, Welch's t-test.

Fig. 4

Expression of GPX as an assessment of Se metabolism in Se-deficient conditions. (A) Experimental scheme for Se-deficient experiments. Assess the reduction of GPX by placing GPX-induced cells in a Se-deficient state. (B) RD cells were treated with selenite 100 nM and SeP 0.5 μg/mL for 24 h. The medium was replaced with Se-depleted medium and further incubated for indicated time period. After that the cells were harvested and WB performed. (C, D) Quantitative values of the (C) GPX1 and (D) GPX4 bands were determined from three independent data sets, corrected for GAPDH, respectively, and shown relative to the SeP 12 h time point as 1. Mean ± S.D., n = 3, SeP vs selenite in each time point, ∗P < 0.05, Welch's t-test followed by Bonferroni correction. (E) SH-SY5Y cells were treated with selenite 1 μM and SeP 5 μg/mL for 24 h. The medium was replaced with Se-depleted medium and further incubated for indicated time period. After that the cells were harvested and WB performed. (F, G) Quantitative values of the (F) GPX1 and (G) GPX4 bands were determined from three independent data sets, corrected for GAPDH, respectively, and shown relative to the SeP 48 h time point as 1. Mean ± S.D., n = 3, SeP vs selenite in each time point, ∗P < 0.05, Welch's t-test followed by Bonferroni correction.

Fig. 5

Contribution of SeP in the defense against ferroptosis in Se deficient states. (A) Experimental scheme to evaluate ferroptosis resistance in Se-deficient conditions. (B, C, D) RD cells were seeded in 96 well plates and treated with selenite 100 nM and SeP 0.5 μg/mL for 24 h. After that, the cells were treated with Se-deficient medium containing stressors, (B) Cumenehydroperoxide (Cum-OOH), (C) Erastin and (D) RSL3 for 24 h. The cell viability was determined using alamarBlue assay. Mean ± S.D., n = 3, ∗P < 0.05, ∗∗P < 0.01, Tukey's multiple comparisons test. (E) Summary of this study. SeP taken up via ApoER2 are degraded and utilized via storage vesicles and are involved in the retention of GPX induction for long periods even under Se-deficient conditions, and repress oxidative stresses e.g., ferroptosis.