Sphingosine-1-phosphate restores endothelial barrier integrity in ovarian hyperstimulation syndrome

Study question: Are follicular fluid (FF) sphingosine-1-phosphate (S1P) levels in patients at risk of developing ovarian hyperstimulation syndrome (OHSS) altered and in part responsible for the high vascular permeability observed in these patients.

Study answer: FF S1P levels are lower in FF from patients at risk of OHSS and treatment with S1P may reduce vascular permeability in these patients.

What is known already: Although advances have been made in the diagnosis, and management of OHSS and in basic knowledge of its development, complete prevention has proven difficult.

Study design, size, duration: A total of 40 FF aspirates were collected from patients undergoing ART. The women (aged 25-39 years old) were classified into a control group (n = 20) or a group at risk of OHSS (n = 20). The EA.hy926 endothelial cell line was used to assess the efffects of FF from patients at risk of OHSS with or without the addition of S1P. An animal model that develops OHSS in immature Sprague-Dawley rats were also used.

Participants/materials, setting, methods: Migration assays, confocal microscopy analysis of actin filaments, immunoblotting and quail chorioallantoic membrane (CAM) assays of in-vivo angiogenesis were performed and statistical comparisons between groups were made.

Main results and the role of chance: The S1P concentration was significantly lower in FF from patients at risk of OHSS (P = 0.03). The addition of S1P to this FF decreased cell migration (P < 0.05) and prevented VE-cadherin phosphorylation in endothelial cells (P < 0.05). S1P in the FF from patients at risk of OHSS increased the levels of VE-cadherin (P < 0.05), N-cadherin (P < 0.05) and β-catenin (P < 0.05), and partially reversed actin redistribution in endothelial cells. The addition of S1P in FF from patients at risk of OHSS also decreased the levels of vascular endothelial growth factor (VEGF₁₂₁; P < 0.01) and S1P lyase (SPL; P < 0.05) and increased the levels of S1PR1 (P < 0.05) in endothelial cells. In CAMs incubated with FF from patients at risk of OHSS with S1P, the number of vessel branch points decreased while the periendothelial cell coverage increased. Additionally, in a rat OHSS model, we demonstrated that vascular permeability and VEGF₁₂₁ and its receptor KDR expression were increased in the OHSS group compared to the control group and that S1P administration decreased these parameters.

Large scale data: N/A.

Limitations, reasons for caution: The results of this study were generated from an in-vitro system. This model reflects the microvasculature in vivo. Even though the ideal model would be the use of human endothelial cells from the ovary, it is obviously not possible to carry out this kind of approach in ovaries of patients from ART. More studies will be necessary to delineate the effects of S1P in the pathogenesis of OHSS. Hence, clinical studies are needed in order to choose the most appropriate method of prevention and management.

Wider implications of the findings: The use of bioactive sphingolipid metabolites may contribute to finding better and safer therapeutic strategies for the treatment of OHSS and other human diseases that display aberrant vascular leakage.

Study funding/competing interests: This work was supported by grants ANPCyT (PICT 2012-897), CONICET (PIP 5471), Roemmers and Baron Foundation, Argentina. The authors declare no conflict of interest.