Serotonin and serotonin reuptake inhibitors alter placental aromatase

Abstract

Serotonin reuptake inhibitors (SRIs) are currently the main molecules prescribed to pregnant women that suffer from depression. Placental cells are exposed to SRIs via maternal blood, and we have previously shown that SRIs alter feto-placental steroidogenesis in an in vitro co-culture model. More specifically, serotonin (5-HT) regulates the estrogen biosynthetic enzyme aromatase (cytochrome P450 19; CYP19), which is disrupted by fluoxetine and its active metabolite norfluoxetine in BeWo choriocarcinoma cells. Based on molecular simulations, the present study illustrates that the SRIs fluoxetine, norfluoxetine, paroxetine, sertraline, citalopram and venlafaxine exhibit binding affinity for the active-site pocket of CYP19, suggesting potential competitive inhibition. Using BeWo cells and primary villous trophoblast cells isolated from normal term placentas, we compared the effects of the SRIs on CYP19 activity. We observed that paroxetine and sertraline induce aromatase activity in BeWo cells, while venlafaxine, fluoxetine, paroxetine and sertraline decrease aromatase activity in primary villous trophoblast. The effects of paroxetine and sertraline in primary villous trophoblasts were observed at the lower doses tested. We also showed that 5-HT and the 5-HT_2A receptor agonist 2,5-dimethoxy-4-iodoamphetamine (DOI) induced CYP19 activity. An increase in phosphorylation of serine and tyrosine and a decrease in threonine phosphorylation of CYP19 was also associated with DOI treatment. Our results contribute to better understanding how 5-HT and SRIs interact with CYP19 and may affect estrogen production. Moreover, this study suggests that alteration of placental 5-HT levels due to depression and/or SRI treatment during pregnancy may be associated with disruption of placental estrogen production.

Keywords: Antidepressant; CYP19; Estrogen; Phosphorylation; Trophoblast.

Fig 1

Molecular docking simulations of serotonin reuptake inhibitors (SRIs) inside the active-site pocket of CYP19 human aromatase. (a) Comparison between the crystallographically resolved aromatase (CYP19)-androstenedione enzyme-substrate complex (PDB 5JL6) and molecular docking predictions. The active-site pocket is depicted as a gold transparent surface, with crystallographically resolved and docked androstenedione substrates shown as green and purple sticks, respectively. The heme cofactor is shown as a ball-and-stick model. (b) Comparative binding energies (kcal*mol⁻¹) for SRIs and androstenedione to CYP19. Binding energies are shown for the three most energetically favorable complexes obtained by virtual docking. Right panel: wire models showing how distinct SRIs can be accommodated in the active-site pocket of CYP19. The crystalographically resolved androstenedione substrate is shown in green. UniProt-defined CYP19 binding site residues D309 and M374 are highlighted by asterisks (UniProt entry P11511)

Fig 2

Effects of serotonin reuptake inhibitors on aromatase activity in BeWo and primary villous trophoblasts. Relative aromatase (CYP19) activity in BeWo cells treated 24 h with (a) citalopram, (b) venlafaxine, (c) paroxetine and (d) sertraline (0.03, 0.1, 0.3 and 1 μM) and in in primary villous trophoblast cells treated 24 h with (e) citalopram, (f) venlafaxine, (g) fluoxetine, (h) paroxetine, (i) sertraline and (j) norfluoxetine (0.03, 0.1, 0.3, 1 and 3 μM). Dashed line represents DMSO-vehicle control. Activities are expressed as a percentage of vehicle-control (DMSO, mean ± SEM). Significant differences compared to control are indicated with asterisks (* P < 0.05; **P < 0.005; ***P < 0.001) as determined by a Kruskal Wallis and Dunn’s post-hoc test, n=3-5

Fig 3

Effect of serotonin and 2,5-dimethoxy-4-iodoamphetamine (DOI) on aromatase in villous trophoblasts. Relative aromatase (CYP19) activity in primary villous trophoblast cells were treated for 24h with (a) serotonin or (b) DOI (5-HT_2A agonist) at different time points. CYP19 activity is expressed in percentage of vehicle-control ± SEM. Protein expression of CYP19 in placental explants treated for 24h with (c) serotonin or (d) 2,5-dimethoxy-4-iodoamphetamine (DOI, 10 and 30 μM). CYP19 protein expression is expressed relative to vehicle-control ± SEM. Dashed line represents DMSO-vehicle control. Significant differences are indicated with asterisks (* P < 0.05; ** P < 0.01). Kruskal Wallis and Dunn’s post-hoc test, n ≥ 3. (e) Representative blot of CYP19 (55 kDa) with total protein used to normalize protein expression. (f) Representative blots (explants isolated from one placenta) of phosphorylated tyrosine, serine, threonine residues and from total protein (at CYP19 molecular weight, 55 kDa) from immunoprecipitated CYP19 in placental explants treated or not with DOI (30 μM) for 15 min, 30 min, 60 min or 24 h.

Fig 4

Predicted conformational changes induced by single-site phosphorylation of human CYP19 aromatase. (a) Single-site phosphorylation of seven target residues predicted in human CYP19 aromatase (see Experimental procedures for details). The active-site pocket is depicted as a gold surface and the phosphate groups are colored as red-orange spheres. (b) Summary of the most important stabilizing interactions of the CYP19-heme-androstenedione ternary complex, as presented according to the relative individual binding energy contribution of each residue. Residues are sorted based on decreasing order of stabilizing effects on the heme cofactor (top 8 residues) or androstenedione substrate (bottom 9 residues). Since the heme group acts as an important energy contributor to the androstenedione substrate, its comparative binding energy contribution is also listed (−33.9 kcal*mol⁻¹). (c) Energy minimization of the ternary complex showing residue conformational changes induced upon single-site phosphorylation events in CYP19. The heme cofactor is represented as a sphere model with carbon, nitrogen, iron and oxygen atoms shown in gray, blue, orange and red, respectively. The androstenedione substrate is shown as green sticks. Residues affected by phosphorylation are shown as blue (unphosphorylated) and gold (phosphorylated) sticks. Relevant residues are labeled according to single letter code numbering