Molecular analysis of nitric oxide production by estrogen and selective estrogen receptor modulators (SERM)

Magda Heras1,2, Ana Paula Dantas1,2. 1Institut d\'Investigacions Biomediques August Pi i Sunyer (IDIBAPS), Cardiology, 08036, Spain, 2Institut Clinic del Torax, 08036, Spain

Estrogen effects are mediated by complex array of signalling pathways that are activated by three distinct receptors: the classical ERα and ERβ and the recently described GPR30. Through some of these pathways, estrogens have been described to promote cardiovascular protection, as for example by increasing the production of nitric oxide. In addition to estrogens, a large number of women are being exposed to selective estrogen receptor modulators (SERM), such as tamoxifen and raloxifene, as a safer alternative to treat the complications of menopause. Although few studies suggest that SERMs display similar protective effects to the cardiovascular system as estrogen, the signaling pathways by which SERMs modulate endothelial function remains unclear. In this study we performed a molecular analysis of the mechanisms trigged by estrogen and the SERMs raloxifene and tamoxifen to modulate nitric oxide (NO) production. Human endothelial cells (HUVEC, Invitrogen) were cultured in phenol-red free DMEM-F12 media and followed the experimental protocols previously described (Hypertension. 2010;56(3):405-11). Cells were treated acutely (5, 15 and 30 minutes) or chronically (48h) with increasing concentrations of estrogen and SERMs (1nM to 1µM). Chronic treatment with estrogen and raloxifene caused a dose-dependent, saturable increase in green fluorescence by DAF-2, indicative of a rise in intracellular NO. Nonetheless maximum fold increase on NO production by raloxifene was significantly lower in comparison to estrogen (Estrogen: 3.2±0.1; Raloxifene: 2.3±0.03, p<0.05). Tamoxifen had no effect on NO production. Raloxifene modulation of eNOS mRNA expression was 60% lower (p < 0.05) than that of estrogen. Estrogen- and raloxifene-induced ERα- and ERβ-mediated activation of eNOS promoter was measured by ChIP and reporter gene assay. By these studies we found that raloxifene display a delayed and weaker interaction of ERα with DNA sequences of eNOS promoter than does estrogen. Besides, raloxifene activity on ERα-mediated activation of eNOS promoter was about 50% lower than that of estrogen, effects that could explain the lower eNOS expression but does not agree with the slighter lower NO production by raloxifene. In this regard, acute effects of raloxifene and estrogen on eNOS phosphorylation at serine 1177 site were evaluated. We found a similar time course and concentration-dependent ability to increase eNOS phosphorylation by both estrogen and raloxifene (Max Fold Increase: estrogen: 2.5±0.1; raloxifene: 2.3±0.3). Estrogen-induced eNOS phosphorylation was attenuated by pre-treating cells (30 min) with 10µM ICI182780 (ERα and ERβ antagonist), 10µM G15 (GPR30 antagonist) and by the phosphoinositide 3-kinase (PI3-K) inhibitor wortmannin (1 µM). By contrast, raloxifene response was not affected by inhibition with ICI182780, suggesting that eNOS phosphorylation by raloxifene is mostly mediated by GPR30. Our finding indicated that the SERM raloxifene is effective to modulate NO production, albeit with a lower potency than estrogen. The differences in the potential of response by estrogen and raloxifene can be explained by the differences in the activation of eNOS expression and activity by those agonists. While estrogen increases both eNOS transcription and phosphorylation mostly via ERα activation, raloxifene display a weaker activation of ERα-mediated eNOS expression and mediates eNOS phosphorylation by a GPR30 and PI3-kinase/Akt mediated signaling pathway.