Vitamin D attenuates sphingosine-1-phosphate (S1P)-mediated inhibition of extravillous trophoblast migration

Abstract

Introduction: Failure of trophoblast invasion and remodelling of maternal blood vessels leads to the pregnancy complication pre-eclampsia (PE). In other systems, the sphingolipid, sphingosine-1-phosphate (S1P), controls cell migration therefore this study determined its effect on extravillous trophoblast (EVT) function.

Methods: A transwell migration system was used to assess the behaviour of three trophoblast cell lines, Swan-71, SGHPL-4, and JEG3, and primary human trophoblasts in the presence or absence of S1P, S1P pathway inhibitors and 1,25(OH)₂D₃. QPCR and immunolocalisation were used to demonstrate EVT S1P receptor expression.

Results: EVTs express S1P receptors 1, 2 and 3. S1P inhibited EVT migration. This effect was abolished in the presence of the specific S1PR2 inhibitor, JTE-013 (p < 0.05 versus S1P alone) whereas treatment with the S1R1/3 inhibitor, FTY720, had no effect. In other cell types S1PR2 is regulated by vitamin D; here we found that treatment with 1,25(OH)₂D₃ for 48 or 72 h reduces S1PR2 (4-fold; <0.05), but not R1 and R3, expression. Moreover, S1P did not inhibit the migration of cells exposed to 1,25(OH)₂D₃ (p < 0.05).

Discussion: This study demonstrates that although EVT express three S1P receptor isoforms, S1P predominantly signals through S1PR2/Gα_12/13 to activate Rho and thereby acts as potent inhibitor of EVT migration. Importantly, expression of S1PR2, and therefore S1P function, can be down-regulated by vitamin D. Our data suggest that vitamin D deficiency, which is known to be associated with PE, may contribute to the impaired trophoblast migration that underlies this condition.

Keywords: GPCR; Lipid signalling; Migration; Placenta; Pre-eclampsia; S1P; Trophoblast; Vitamin D.

Fig. 1

S1PR1-3 G-protein linked signalling cascades.

Fig. 2

Expression of S1P receptors in primary extravillous trophoblast (EVT) and EVT cell lines. S1PR protein localisation was determined by immunostaining of human first trimester placenta (A), Swan-71 cells (B) and SGHPL-4 cells (C) counterstained with hematoxylin or DAPI. Staining for HLA-G was used to indicate extravillous trophoblast and some slides were incubated with IgG instead of primary antibody as a negative control. Bar represents 20 μM. Images are representative of 4 independent experiments.

Fig. 3

S1P attenuates EVT migration via S1PR2. (A) Swan-71 cells were seeded on cell culture transwell inserts, incubated for 24 h with varying concentrations of S1P (100nM-10μM) or vehicle control (methanol) and then the number of cells that had migrated through the filter (8 μM pores) was determined by counting the number of cells present in micrographs taken, using an Olympus inverted microscope at 10× magnification, of six different fields for each insert. Data are presented as percentage of cells, relative to control, that had migrated. Bar – mean ± SEM; § - p < 0.05; n = 5. (B) Primary placental villous tips were embedded in collagen, cultured for 24 h and then photographed (baseline). Tissue was treated with media containing either 1% fetal calf serum (control), 100 nM S1P or 100 nM S1P and 100 nM JTE013 for a further 24 h before photographing the same area of explant. The images shown are representative of those obtained from four independent experiments. Swan-71 cells (C), SGHPL-4 cells (D) or JEG3 cells (E) were incubated with 100 nM of S1PR1/3 inhibitor, FTY720 (C-D) or FTY720p (E) or the S1PR2 inhibitor, JTE-013 (100 nM; C-E) for 30 min before the addition of vehicle or S1P (100 nM) and culture for 24 h. The number of cells that had migrated through the filter is shown as percentage relative to untreated cells. Bar – mean ± SEM; § - p < 0.05 versus untreated cells; n = 7.

Fig. 4

S1PR1 is active in EVT cells. (A) Swan-71 cells were cultured in the absence or presence of forskolin (1 μM) for 15 min. The S1PR2 inhibitor JTE-013 (100 nM), the S1P R1/3 inhibitor FTY720 (100 nM) or vehicle was added for 20 min then cells were cultured in the absence or presence of 100 nM S1P for a further 30 min before measuring intracellular cAMP levels. Data are from four independent experiments and are expressed relative to control. Bar – mean ± SEM; § - p < 0.05 versus untreated cells; $ - p < 0.05 versus forskolin treatment alone; # - p < 0.05 versus forskolin and S1P treatment. (B) Swan-71 cells were exposed to S1P (0-100 nM) for 10 min or 1 h then cell lysates were analysed by western blotting using antibodies that recognise phosphorylated (pAkT) or total Akt. Blots are representative of data obtained from 3 independent experiments.

Fig. 5

S1P activates RhoA/ROCK but inhibits Rac. Swan-71 cells were treated with vehicle (control), S1P (100 nM) or the Rho kinase inhibitor Y-27632 (10 μM) for 30 min followed by S1P (100 nM) for a further 30 min or 24 h before staining with Alexa Fluor 594-conjugated phalloidin and DAPI (A) or assessing migration (B) respectively. Bar – mean ± SEM; n = 6; § - p < 0.05 relative to culture in the absence of S1P. (C) Swan-71 cells were cultured in the absence (control) or presence of S1P (100 nM) or a known stimulator of cAMP synthesis, EGF (100 ng/ml), for 60 s, lysed and then the level of active Rac in 0.5 mg/ml protein from each sample was determined using the G-LISA Rac Activation Biochem Kit. Bar – mean ± SEM; n = 5; § - p < 0.05 relative to control.

Fig. 6

Vitamin D suppresses S1PR2 expression and attenuates S1P-mediated inhibition of EVT cell migration. Swan-71 cells were incubated with vehicle or 1-10 nM 1,25-dihydroxyvitamin D (D₃) for 48 or 72 h then (A) the level of S1PR2 mRNA, normalised to the housekeeping gene YWHAZ, was measured by qPCR (n = 3). (B) Swan-71, SGHPL-4 or JEG3 cells were cultured in 10 nM D₃ for 48 h then seeded in transwells and incubated in the absence or presence of 100 nM S1P for a further 24 h before assessing migration (n = 5-7). Bar – median; § - p < 0.05 versus cells cultured in the relevant control medium (absence or presence of D₃), normalised to 100%.