Selenium Prevents Inflammation in Human Placenta and Adipose Tissue In Vitro: Implications for Metabolic Diseases of Pregnancy Associated with Inflammation

Abstract

Gestational diabetes mellitus (GDM) and maternal obesity are significant metabolic complications increasingly prevalent in pregnancy. Of major concern, both GDM and maternal obesity can have long-term detrimental impacts on the health of both mother and offspring. Recent research has shown that increased inflammation and oxidative stress are two features central to the pathophysiology of these metabolic conditions. Evidence suggests selenium supplementation may be linked to disease prevention in pregnancy; however, the specific effects of selenium on inflammation and oxidative stress associated with GDM and maternal obesity are unknown. Therefore, this study aimed to investigate the effect of selenium supplementation on an in vitro model of GDM and maternal obesity. Human placental tissue, visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) were stimulated with either the bacterial product lipopolysaccharide (LPS) or the pro-inflammatory cytokine TNF-α. Selenium pre-treatment blocked LPS and TNF-α induced mRNA expression and secretion of pro-inflammatory cytokines and chemokines, while increasing anti-inflammatory cytokine and antioxidant mRNA expression in placenta, VAT and SAT. Selenium pre-treatment was also found to inhibit LPS- and TNF-α induced phosphorylation of ERK in placenta, VAT and SAT. These findings indicate that selenium may be able to prevent inflammation and oxidative stress associated with GDM and maternal obesity. Additional in vivo studies are required to identify the efficacy of selenium supplementation in preventing inflammatory pathways activated by GDM and maternal obesity and to elucidate the mechanism involved.

Keywords: adipose tissue; gestational diabetes; inflammation; insulin resistance; maternal obesity; placenta; selenium.

Figure 1

Effect of selenium on pro-inflammatory cytokine expression in human placenta. Human placenta was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,B,D,E) GM-CSF, IL1-A, IL1-B and IL6 mRNA expression were analysed by RT-qPCR and fold change was calculated relative to basal expression. (C,F) The concentration of IL1-A and IL6 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF-α; repeated measures one-way ANOVA.

Figure 2

Effect of selenium on pro-inflammatory cytokine expression in VAT and SAT from pregnant women. (A–E) VAT (n = 6 patients) and (F–J) SAT (n = 6 patients) were incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h. (A–D,F–I) GM-CSF, IL1-A, IL1-B and IL6 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (E,J) The concentration of IL6 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 3

Effect of selenium on anti-inflammatory cytokine expression in human placenta, and VAT and SAT from pregnant women. (A,B) Placenta (n = 6 patients), (C,D) VAT (n = 6 patients) and (E,F) SAT (n = 6 patients) were incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h. IL4 and IL13 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. For all graphs, individual data points represent 6 independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 4

Effect of selenium on CCL chemokine expression in human placenta. Placenta was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,C,D,F) CCL2, CCL3, CCL4 and CCL8 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (B,E) The concentration of CCL2 and CCL4 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 5

Effect of selenium on CXCL chemokine expression in human placenta. Placenta was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,C,D,F,H) CXCL1, CXCL2, CXCL5, CXCL8 and CXCL10 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (B,E,G) The concentration of CXCL1, CXCL5 and CXCL8 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 6

Effect of selenium on CCL chemokine expression in VAT from pregnant women. VAT was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,C,D,F) CCL2, CCL3, CCL4 and CCL8 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (B,E) The concentration of CCL2 and CCL4 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 7

Effect of selenium on CXCL chemokine expression in VAT from pregnant women. VAT was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,C,D,F,H) CXCL1, CXCL2, CXCL5, CXCL8 and CXCL10 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (B,E,G) The concentration of CXCL1, CXCL5 and CXCL8 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 8

Effect of selenium on CCL chemokine expression in SAT from pregnant women. SAT was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,C,D,F) CCL2, CCL3, CCL4 and CCL8 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (B,E) The concentration of CCL2 and CCL4 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 9

Effect of selenium on CXCL chemokine expression in SAT from pregnant women. SAT was incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h (n = 6 patients). (A,C,D,F,H) CXCL1, CXCL2, CXCL5, CXCL8 and CXCL10 mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. (B,E,G) The concentration of CXCL1, CXCL5 and CXCL8 in the conditioned media was assayed by ELISA. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 10

Effect of selenium on antioxidant selenoenzyme expression in human placenta, and VAT and SAT from pregnant women. (A,B) Placenta (n = 6 patients), (C,D) VAT (n = 6 patients) and (E,F) SAT (n = 6 patients) were incubated with or without 10 μM sodium selenite (Sel) for 1 h and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 20 h. GPx and TrxR mRNA expression was analysed by RT-qPCR and fold change was calculated relative to basal expression. For all graphs, individual data points represent six independent experiments and are displayed as mean ± SEM. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.

Figure 11

Effect of selenium on ERK activation in human placenta, and VAT and SAT from pregnant women. (A) Placenta (n = 5 patients), (B) VAT (n = 5 patients) and (C) SAT (n = 5 patients) were incubated with or without 10 μM sodium selenite (Sel) overnight and then stimulated with 10 µg/mL LPS or 10 ng/mL TNF-α for a further 15 min. The protein expression of phosphorylated ERK1 (pERK) and total ERK1 was analysed by Western blotting; pERK protein expression was normalised to total ERK expression. A Representative western blot from one patient is shown. For all graphs, the data were calculated relative to basal expression and displayed as mean ± SEM with data points representing five individual experiments. ^a p ≤ 0.05 vs. LPS, ^b p ≤ 0.05 vs. TNF; repeated measures one-way ANOVA.