Selenate enhances STAT3 transcriptional activity in endothelial cells: differential actions of selenate and selenite on LIF cytokine signaling and cell viability

Abstract

Sodium selenate may have utility in treating Alzheimer's disease and diabetes; however, its impact on the associated proinflammatory cytokine signaling of endothelial cells has not been investigated. We report that treatment of human microvascular endothelial cells with sodium selenate at a pharmacological dose (100 μM) enhanced tyrosine phosphorylation of nuclear STAT3 on Y705 in response to IL-6-type cytokine, leukemia inhibitory factor (LIF), indicative of enhanced STAT3 activity. Accordingly, STAT3 nuclear binding to DNA was increased, as well as LIF-induced gene expression of chemokine (C-C motif) ligand 2 (CCL2). CCL2 plays a key role in inflammatory processes associated with neuronal degenerative and vascular diseases. The enhancing action of selenate on LIF-induced STAT3 Y705 phosphorylation was replicated by vanadate and a specific inhibitor of protein tyrosine phosphatase, non-receptor type 1 (PTP1B). Moreover, we observed that selenite, the cellular reduction bioproduct of selenate but not selenate itself, inhibited enzymatic activity of human recombinant PTP1B. Our findings support the conclusion that in human microvascular endothelial cells selenate has a vanadate-like effect in inhibiting PTP1B and enhancing proinflammatory STAT3 activation. These findings raise the possibility that beneficial actions of supranutritional levels of selenate for treating Alzheimer's and diabetes may be offset by a proinflammatory action on endothelial cells.

Figure 1

Selenate enhances nuclear STAT3 phosphorylation. HMEC-1 were pretreated for 2 h with 100 μM sodium selenate or vehicle and then 2 ng/mL LIF or vehicle was added for 1 h. Nuclear extracts were prepared and levels of (A) pY705 STAT3 and STAT3 or (C) pS727 STAT3 and STAT3 evaluated by Western blotting. (B & D) Results were quantified using the LI-COR Odyssey system and the ratio of phosphorylated STAT3 to total STAT3 determined. Results of individual experiments were normalized to the control. Values are mean ± SEM of 7 independent observations. *P < 0.05; **P < 0.01 (column below line origin vs. column below line end) by ANOVA and Newman-Keuls Multiple Comparison Test.

Figure 2

Selenite does not enhance nuclear STAT3 phosphorylation. HMEC-1 were pretreated for 2 h with 100 μM sodium selenite or vehicle and then 2 ng/mL LIF or vehicle was added for 1 h. (A) Nuclear extracts were prepared and levels of pY705 STAT3 evaluated by Western blotting. (B) Results were quantified using the LI-COR Odyssey system and the ratio of phosphorylated STAT3 to total STAT3 determined. Results of individual experiments were normalized to the control. Values are mean ± SEM of 3 independent observations. ***P < 0.001 vs. control and selenate by ANOVA and Newman-Keuls Multiple Comparison Test. ns = not significantly different.

Figure 3

Effect of selenate or selenite on cell viability. HMEC-1 were treated with vehicle (control), 100 μM sodium selenate, 100 μM sodium selenite, or 100 μM calyculin A for 24 h. The numbers of total and live cells under each condition were determined using a NucleoCounter as described in 2.5 Cell counts and viability. Numbers of total and live cells are expressed as a percent of the respective control (vehicle-treated) number (left y-axis). The fraction of live cells is the ratio LIVE/TOTAL (right y-axis). Values are mean ± SEM of 3 independent observations. *P < 0.05 and **P < 0.001 vs. respective control by ANOVA and Dunnett’s Multiple Comparison Test.

Figure 4

Selenate enhances nuclear STAT3 binding activity. HMEC-1 were pretreated for 2 h with 100 μM sodium selenate or vehicle and then 2 ng/mL LIF or vehicle was added for 1 h. Nuclear extracts were prepared and equal protein amounts assessed for binding to a STAT3 consensus binding motif using a fluorescent ELISA-based assay. Values are mean ± SEM of 7 independent observations. Statistical significance was determined using ANOVA and the Newman-Keuls Multiple Comparison Test. ***P < 0.001 (column below line origin vs. column below line end).

Figure 5

Effect of selenate on LIF-induced gene expression. HMEC-1 were pretreated for 2 h with 100 μM sodium selenate or vehicle and then 2 ng/mL LIF or vehicle was added for 1 h. RNA was extracted, reverse transcribed, and analyzed for SOCS3 and CCL2 expression by real time PCR. Results were normalized to levels of GAPDH and expressed as the fold increase over control levels (vehicle only). Values from 5 independent experiments are shown along with the mean (dotted line) ± SEM. ** P < 0.01 by paired t-test.

Figure 6

Vanadate and PTP1B inhibitor enhance nuclear STAT3 Y705 phosphorylation. (A) Cells were pretreated 2h with 100 μM sodium vanadate or vehicle, and then treated for 1 h with 2 ng/mL LIF or vehicle. Nuclear extracts were prepared and analyzed for levels of STAT3 Y705 phosphorylation and STAT3. Results were quantified by the LI-COR Odyssey Detection System. Values are mean ± SEM of 3 independent observations. (B) Cells were pretreated 1h with 50 μM sodium PTP1B inhibitor or vehicle (0.02% v/v DMSO), and then treated for 1 h with 2 ng/mL LIF or vehicle. Nuclear extracts were prepared and analyzed by Western analysis. Shown is a representative blot. (C) Levels of STAT3 Y705 phosphorylation and STAT3 from (B) were quantified by the LI-COR Odyssey Detection System. Values are mean ± SEM of 4 independent observations. *P < 0.05, **P < 0.01, and ***P < 0.001 by ANOVA and Newman-Keuls Multiple Comparison Test.

Figure 7

Effect of selenite or selenate on PTP1B activity. A colorimetric assay was used to assess the direct effect of sodium selenite or sodium selenate on PTP1B enzymatic activity. The assay measures free phosphate formed from a phosphopeptide sequence based on the insulin receptor β subunit domain. The reaction contained human recombinant 25 ng/mL PTP1B and 75 μM substrate. Sodium selenate and sodium selenite were added at a final concentration of 500 or 5000 μM. Absorbance readings were corrected by subtracting the blank value obtained by incubating PTP1B without substrate. Values are mean ± SEM of 3 independent observations. *P < 0.05 vs. control by ANOVA and Dunnett’s Multiple Comparison Test.