Sphingosine signalling regulates decidual NK cell angiogenic phenotype and trophoblast migration.

STUDY QUESTION: Is sphingosine-1-phosphate (S1P) signalling involved in the regulation of the angiogenic function of decidual (d)NK cells during human pregnancy?

SUMMARY ANSWER: Human dNK cells, characterized by S1P receptor 5 (S1PR5) expression, are reactive to microenvironmental S1P to modify their VEGF expression and to regulate trophoblast migration and endothelial angiogenesis.

WHAT IS KNOWN ALREADY: S1P signalling can modulate peripheral (p)NK cells migration and function. As a unique NK population, human dNK can produce multiple cytokines and angiogenic growth factors to mediate extravillous trophoblast (EVT) invasion and spiral artery remodelling during pregnancy.

STUDY DESIGN, SIZE, DURATION: The study was designed to examine S1PR expression and function by freshly isolated human dNK cells in response to different S1P scenarios, created by FTY720, an S1P analogue and S1PR modulator. Ex vivo and in vitro experiments were performed to evaluate the functions of dNK cells. The study was performed between September 2011 and June 2013.

PARTICIPANTS/MATERIALS, SETTING, METHODS: Human peripheral blood and decidual samples were collected and the S1PR expression by the decidual leukocytes population was examined. FTY720-induced dNK phenotypic and functional changes (including VEGF and IL-8 expression) were evaluated by multi-colour flow cytometric assays and transwell migration studies. Human placental explant culture and wound healing assays were performed to investigate whether S1P-activated dNK mediated trophoblast migration while angiogenesis was assessed by human umbilical vein endothelial cells (HUVEC) tube formation assays. Both first and second trimester dNK cells were studied to compare the difference in S1PR expression over time at the fetal-maternal interface.

MAIN RESULTS AND THE ROLE OF CHANCE: Freshly isolated NK cells (CD45(+)CD56(+)CD16(-)) from blood (pNK) and decidua (dNK) had low S1PR1 reactivity while S1PR5 was prominently expressed by dNK (40%) and, to a lesser extent, by pNK (18%; P < 0.05) cells. S1PR5 expression by dNK was significantly down-regulated by FTY720 treatment, which also impaired decidual leukocyte mobility and cellular contact with invasive EVT. FTY720 significantly reduced VEGF expression by dNK, both in the numbers of VEGF(+) cells and in fluorescence intensity (P < 0.05). IL-8 expression by dNK was not changed by FTY720 and remained low at 8% positivity. Trophoblast migration and HUVEC tube formation were stimulated by control leukocytes, enriched CD56(+) dNK or their conditioned medium, respectively, but this effect was markedly abrogated once they were pretreated with FTY720 (P < 0.05). There was a significant decrease in S1PR5 expression in second trimester dNK cells, compared with those from first trimester (P < 0.05). No significant differences in the levels of angiogenic factors (VEGF or IL-8) were detected between first and second trimester dNK cells.

LIMITATIONS, REASONS FOR CAUTION: Our ex vivo and in vitro experimental samples were from healthy women undergoing elective pregnancy termination. FTY720 is a chemical ligand for the S1PRs; little is known regarding the levels or actions of the naturally occurring ligand S1P in human gestational tissues. The in vivo function of S1PR5(+) dNK may be further investigated by using a genetically modified animal model.

WIDER IMPLICATIONS OF THE FINDINGS: This is the first study to investigate the role of S1PR and S1P interaction on dNK cell physiology and their downstream effects on trophoblast migration. We suggest that S1PR5 may represent a potential target for cellular targeted treatments for gestational diseases such as pre-eclampsia and intrauterine growth restriction that are characterized by inadequate dNK/trophoblast-coordinated uterine spiral artery transformation.

STUDY FUNDING/COMPETING INTEREST(S): This study was supported by Canadian Institutes of Health Research (CIHR), MOP82811 to Dr S.J.L.