Advertisement

Plasma S1P and sphingosine are not different prior to preeclampsia in women at high risk of developing the disease

  • Edward D. Johnstone
    Correspondence
    Corresponding author:
    Affiliations
    Maternal and Fetal Health Research Centre, School of Medical Sciences, , The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
  • Melissa Westwood
    Affiliations
    Maternal and Fetal Health Research Centre, School of Medical Sciences, , The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
  • Mark Dilworth
    Affiliations
    Maternal and Fetal Health Research Centre, School of Medical Sciences, , The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
  • Jonathan R. Wray
    Affiliations
    Laboratory for Lipidomics and Lipid Biology, Division of Pharmacy and Optometry, School of Health Sciences, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
  • Alexandra C. Kendall
    Affiliations
    Laboratory for Lipidomics and Lipid Biology, Division of Pharmacy and Optometry, School of Health Sciences, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
  • Anna Nicolaou
    Affiliations
    Laboratory for Lipidomics and Lipid Biology, Division of Pharmacy and Optometry, School of Health Sciences, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK

    Lydia Becker Institute of Immunology and Inflammation, Faculty of Biology, Medicine and Health, The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
  • Jenny E. Myers
    Affiliations
    Maternal and Fetal Health Research Centre, School of Medical Sciences, , The University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
    Search for articles by this author
Open AccessPublished:November 09, 2022DOI:https://doi.org/10.1016/j.jlr.2022.100312

      Abstract

      Introduction

      Sphingolipids like sphingosine-1-phosphate (S1P) have been implicated in the pathophysiology of preeclampsia. We hypothesised that plasma S1P would be increased in women at high risk of developing preeclampsia who subsequently develop the disease. Low circulating placental growth factor (PlGF) is known to be associated with development of preeclampsia so further we hypothesised that increased S1P would be associated with concurrently low PlGF.

      Materials and Methods

      Case control study using stored maternal blood samples from 14-24 weeks pregnancy, collected from 95 women at increased risk of preeclampsia. Pregnancy outcome was classified as uncomplicated, preterm preeclampsia (<37 weeks) or term preeclampsia. Plasma lipids were extracted and analysed by ultraperformance liquid chromatography coupled to electrospray ionisation tandem mass spectrometry to determine concentrations of S1P and sphingosine.

      Results

      Median plasma S1P was 0.339nmol/ml and sphingosine was 6.77nmol/l. There were no differences in the plasma concentrations of S1P or sphingosine in women who subsequently developed preeclampsia, no effect of gestational age, fetal sex, ethnicity, or presence of pre-existing hypertension. There was a correlation between S1P and sphingosine plasma concentration (p<0.0001). There was no relationship between S1P or sphingosine with PlGF.

      Discussion

      Previous studies have suggested that plasma S1P maybe a biomarker of preeclampsia. In our larger study we failed to demonstrate this is women at high risk of developing the disease. We did not show a relationship with known biomarkers of the disease, suggesting that S1P is unlikely to be a useful predictor of the development of preeclampsia later in pregnancy.

      Key words

      Introduction

      Sphingolipids affect a diverse array of cellular processes and functions in multiple biological systems. Deranged signalling has been linked to multiple diseases (
      • Huang W.C.
      • Liang J.
      • Nagahashi M.
      • Avni D.
      • Yamada A.
      • Maceyka M.
      • et al.
      Sphingosine-1-phosphate phosphatase 2 promotes disruption of mucosal integrity, and contributes to ulcerative colitis in mice and humans.
      ). Sphingosine-1-Phosphate (S1P) is a pleiotropic sphingolipid found in the circulation that has been extensively investigated in the context of multiple pathologies in part due to its potent effects on cell motility (
      • Pyne S.
      • Lee S.C.
      • Long J.
      • Pyne N.J.
      Role of sphingosine kinases and lipid phosphate phosphatases in regulating spatial sphingosine 1-phosphate signalling in health and disease.
      ). Circulating S1P levels are tightly controlled by sphingosine kinase 1 and 2 , S1P phosphatases (S1PP)(
      • Huang W.C.
      • Liang J.
      • Nagahashi M.
      • Avni D.
      • Yamada A.
      • Maceyka M.
      • et al.
      Sphingosine-1-phosphate phosphatase 2 promotes disruption of mucosal integrity, and contributes to ulcerative colitis in mice and humans.
      ), lipid phosphate phosphatases (
      • Pyne S.
      • Lee S.C.
      • Long J.
      • Pyne N.J.
      Role of sphingosine kinases and lipid phosphate phosphatases in regulating spatial sphingosine 1-phosphate signalling in health and disease.
      ) and S1P lyase which degrades it into sphingosine and ethanolamine as part of the so-called sphingolipid rheostat (
      • Newton J.
      • Lima S.
      • Maceyka M.
      • Spiegel S.
      Revisiting the sphingolipid rheostat: Evolving concepts in cancer therapy.
      ). S1P affects cell movement via interactions with five G-protein linked cell surface receptors (S1PR1-5) and in vitro has been shown to affect the migratory cells of the placenta, extra-villous trophoblast (EVT), by interaction with S1PR2 (
      • Westwood M.
      • Al-Saghir K.
      • Finn-Sell S.
      • Tan C.
      • Cowley E.
      • Berneau S.
      • et al.
      Vitamin D attenuates sphingosine-1-phosphate (S1P)-mediated inhibition of extravillous trophoblast migration.
      ). Sphingosine kinase deficient mice demonstrate impaired placental development (
      • Mizugishi K.
      • Li C.
      • Olivera A.
      • Bielawski J.
      • Bielawska A.
      • Deng C.X.
      • et al.
      Maternal disturbance in activated sphingolipid metabolism causes pregnancy loss in mice.
      ). Impaired placental development and impaired EVT migration have been implicated in the pathophysiology of preeclampsia which affects ∼3% of pregnancies and causes significant maternal and infant morbidity and mortality worldwide. S1P is potential mediator of the disease. S1P concentrations may also be altered in inflammatory states (
      • Nagahashi M.
      • Abe M.
      • Sakimura K.
      • Takabe K.
      • Wakai T.
      The role of sphingosine-1-phosphate in inflammation and cancer progression.
      ) further supporting its involvement in the disease, as preeclampsia is also, in part, an inflammatory disorder (
      • Rana S.
      • Lemoine E.
      • Granger J.P.
      • Karumanchi S.A.
      Preeclampsia: Pathophysiology, Challenges, and Perspectives.
      ). Sphingosine is present at a much lower fraction in plasma than S1P (
      • Hammad S.M.
      • Al Gadban M.M.
      • Semler A.J.
      • Klein R.L.
      Sphingosine 1-phosphate distribution in human plasma: associations with lipid profiles.
      ) and little is known about control of plasma concentrations.
      Studies using HPLC methods in non-pregnant healthy adult populations of men and women have demonstrated plasma S1P levels of 0.75 ± 0.16 nmol/mL (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ) to 1nmol/ml (
      • Halilbasic E.
      • Fuerst E.
      • Heiden D.
      • Japtok L.
      • Diesner S.C.
      • Trauner M.
      • et al.
      Plasma Levels of the Bioactive Sphingolipid Metabolite S1P in Adult Cystic Fibrosis Patients: Potential Target for Immunonutrition?.
      ). Concentrations increase in inflammatory and infectious disease as part of the body’s immune response (
      • Hsu S.C.
      • Chang J.H.
      • Hsu Y.P.
      • Bai K.J.
      • Huang S.K.
      • Hsu C.W.
      Circulating sphingosine-1-phosphate as a prognostic biomarker for community-acquired pneumonia.
      ,
      • Obinata H.
      • Hla T.
      Sphingosine 1-phosphate and inflammation.
      ). In pregnant women studies have reported conflicting data on S1P concentrations in normal pregnancy and diseases associated with placental dysfunction. Melland-Smith et al. (
      • Melland-Smith M.
      • Ermini L.
      • Chauvin S.
      • Craig-Barnes H.
      • Tagliaferro A.
      • Todros T.
      • et al.
      Disruption of sphingolipid metabolism augments ceramide-induced autophagy in preeclampsia.
      ) demonstrated reduced serum S1P in women with preeclampsia versus normal pregnancy. This contrasted with findings by Dobierzewska et al. (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ) who showed no significant difference in plasma S1P, at any gestation, in women who went on to develop preeclampsia versus controls. These studies are both small (n=10 and 7 respectively) and contrasting results may be explained by their potentially unrepresentative sample size; the Melland-Smith study differed in that the investigators analysed samples taken nearer to diagnosis of preeclampsia. A larger study of 57 women with a less well-defined preeclampsia phenotype suggested an increase in plasma S1P in women with diagnosed preeclampsia (
      • Charkiewicz K.
      • Goscik J.
      • Blachnio-Zabielska A.
      • Raba G.
      • Sakowicz A.
      • Kalinka J.
      • et al.
      Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
      ).
      Circulating levels of sphingosine are 8.0nmol/l in non-pregnant subjects (
      • Hammad S.M.
      • Al Gadban M.M.
      • Semler A.J.
      • Klein R.L.
      Sphingosine 1-phosphate distribution in human plasma: associations with lipid profiles.
      ). To our knowledge there is only one previous study examining plasma sphingosine concentration in preeclampsia and this suggested that the plasma concentrations were significantly higher than in the non-pregnant population (
      • Charkiewicz K.
      • Goscik J.
      • Blachnio-Zabielska A.
      • Raba G.
      • Sakowicz A.
      • Kalinka J.
      • et al.
      Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
      ).
      In this study we examined second and early third trimester S1P and sphingosine plasma concentrations in a case-control study which included women at high risk of developing preeclampsia. All women included were at increased risk of preeclampsia either because of a history of a hypertensive pregnancy disorder and/or because of chronic pre-pregnancy hypertension. Based on the observed inhibitory effects of S1P on EVT migration in vitro (
      • Westwood M.
      • Al-Saghir K.
      • Finn-Sell S.
      • Tan C.
      • Cowley E.
      • Berneau S.
      • et al.
      Vitamin D attenuates sphingosine-1-phosphate (S1P)-mediated inhibition of extravillous trophoblast migration.
      ) we hypothesised that S1P concentrations would be increased in women that went on to develop preeclampsia. We investigated the relationship between measurements of S1P, sphingosine and the preeclampsia biomarker, placental growth factor (PlGF), and compared different phenotypes of preeclampsia (preterm and term). In a subset of women with longitudinal samples, we compared measurements between the second and early third trimester.

      Materials and Methods

      This case control study used plasma samples collected from the Manchester Antenatal Vascular Service (The MAViS clinic (
      • Shawkat E.
      • Mistry H.
      • Chmiel C.
      • Webster L.
      • Chappell L.
      • Johnstone E.D.
      • et al.
      The effect of labetalol and nifedipine MR on blood pressure in women with chronic hypertension in pregnancy.
      )). Collection of samples was approved by the NRES Committee North West 11/NW/0426 in accordance with the Declaration of Helsinki principles; all women gave written informed consent to donate samples for research studies. Women recruited had either current hypertension or a hypertensive disorder in a prior pregnancy. Clinical data and blood samples in EDTA were collected during routine visits between 14-17+6 and 18+0-24+6 weeks; for a small number of women longitudinal samples were available. For women included in this case-control study, pregnancy outcomes were categorised as uncomplicated (birth ≥37 weeks, birthweight centile ≥10th with no hypertensive complications) or preeclampsia (defined using ISSHP (
      • Brown M.A.
      • Magee L.A.
      • Kenny L.C.
      • Karumanchi S.A.
      • McCarthy F.P.
      • Saito S.
      • et al.
      The hypertensive disorders of pregnancy: ISSHP classification, diagnosis & management recommendations for international practice.
      ) guidelines). Women with preeclampsia were divided into preterm preeclampsia (women who developed the disease before 37 weeks and who were delivered by this gestation) and term preeclampsia (women who developed preeclampsia after 37 weeks and were delivered after this gestation). In women with pre-pregnancy hypertension, pre-eclampsia was defined as worsening hypertension associated with evidence of placental dysfunction, proteinuria and/or multiorgan disease. All samples were taken prior to a diagnosis of preeclampsia. There were 127 measurements in total in 95 women; 65 women has a single measurement and 30 women had two measurements.

      Sphingolipid extraction and UPLC-MS/MS analysis

      Lipid extractions were carried out as previously described (
      • Kendall A.C.
      • Pilkington S.M.
      • Massey K.A.
      • Sassano G.
      • Rhodes L.E.
      • Nicolaou A.
      Distribution of bioactive lipid mediators in human skin.
      ) using a single-phase system to maximise recovery(
      • Bielawski J.
      • Szulc Z.M.
      • Hannun Y.A.
      • Bielawska A.
      Simultaneous quantitative analysis of bioactive sphingolipids by high-performance liquid chromatography-tandem mass spectrometry.
      ). Briefly, plasma (50 μl) was added to ice-cold ethyl acetate: isopropanol: water (6:3:1, v/v/v) and spiked with 4 ng each of deuterated internal standards for sphingosine (sphingosine-d7; Avanti Polar Lipids, Alabaster, AL, USA) and S1P (S1P-d7; Avanti Polar Lipids). Samples were then incubated on ice for 30 min, centrifuged to pellet out the denatured proteins, and the supernatant was collected and dried down under a gentle stream of nitrogen. The resulting lipid residues were resuspended in methanol containing 0.1 % (v/v) formic acid (mobile phase B) and stored at -20˚C until analysis. Assay recoveries for sphingosine-d7 and S1P-d7 were 87.7 % and 90.1 %, respectively. Sphingosine and S1P were analysed by ultraperformance liquid chromatography couple to electrospray ionisation tandem mass spectrometry (UPLC/ESI-MS/MS), as previously described (
      • Kendall A.C.
      • Kiezel-Tsugunova M.
      • Brownbridge L.C.
      • Harwood J.L.
      • Nicolaou A.
      Lipid functions in skin: Differential effects of n-3 polyunsaturated fatty acids on cutaneous ceramides, in a human skin organ culture model.
      ), using an Acquity UPLC system (Waters Corporation, Wilmslow, UK) paired with a triple quadrupole mass spectrometer (Xevo TQ-S; Waters Corporation); data acquisition using MassLynx software (Waters Corporation). Separation was performed using a reverse-phase Acquity BEH C8 column (1.7 μm 2.1 x 100mm) at a flow rate of 0.3 ml/min and a column temperature of 30˚C. Separation was performed using a gradient system of mobile phase A (water containing 0.1 % (v/v) formic acid) and mobile phase B (methanol containing 0.1 % (v/v) formic acid). Electrospray ionisation was performed in positive-ion mode using the following settings: capillary voltage, 3.5kV; source temperature, 100˚C; desolvation gas temperature, 450˚C. Analytes were quantitated using multiple reaction monitoring (MRM): S1P m/z 380.249>264.278; S1P-d7 m/z 387.292>271.25; sphingosine m/z 300.332>282.205; sphingosine-d7 m/z 307.300>289.300.
      Calibration lines were constructed using synthetic standards (Avanti Polar Lipids) to allow accurate quantitation. Results are expressed as ng/mL plasma and have been converted to nmol/ml to aid comparison with previously published results.

      PlGF assay

      PlGF levels were obtained using the Roche Elecsys automated platform as previously described (
      • Verlohren S.
      • Herraiz I.
      • Lapaire O.
      • Schlembach D.
      • Zeisler H.
      • Calda P.
      • et al.
      New gestational phase-specific cutoff values for the use of the soluble fms-like tyrosine kinase-1/placental growth factor ratio as a diagnostic test for preeclampsia.
      ). Published centile distribution reference ranges were used to determine <5th centile within our dataset.

      Data processing and statistics

      Data analysis was performed using GraphPad Prism version 8 for Windows (GraphPad Software, La Jolla, CA, USA) and STATA v13 (StataCorp LLC4905 Lakeway Drive, College Station, Texas). Distribution of data was checked for normality and logged where appropriate. Single group comparisons were made using non-parametric tests (Mann-Whitney or Kruskall-Wallis). Correlation between continuous measurements was assessed using linear regression; the earliest measurement per woman was included in both the regression and between group analyses. The impact of subsequent preeclampsia on the relationship between analytes was assessed using an interaction term. P< 0.05 was considered significant for all analyses.

      Results

      Participant characteristics are shown in Table 1. There were no differences between women with different pregnancy outcomes. Most women included in the study had pre-pregnancy hypertension. There were slightly more male infants in the preterm preeclampsia group and women with preterm preeclampsia delivered earlier and with lower birthweight infants as expected. To determine if there were potential confounding demographic variables within the dataset that were masking any differences between outcome groups, we compared S1P and sphingosine levels between difference ethnicities (supp Fig. 1), between male and female fetuses (supp Fig.2) and women with and without pre-existing hypertension (supp Fig. 3). There were no differences between any compared groups. Linear regression did not identify a significant relationship between either S1P or sphingosine and birthweight (p=0.987 and 0.827 respectively).
      Table 1Demographics of women included in the study.
      Uncomplicated pregnancy outcome n=52 (%)Preterm preeclampsia n=27 (%)Term preeclampsia n=16 (%)P value
      Maternal Age (Median, IQR)34.6, 5.532.4, 8.433.3, 5.50.054
      Ethnicity (white)25 (48)15 (55)9 (56)0.753
      Gestational age at delivery days (Median, IQR)269 (266-274.5)232 (214-245)265 (263.5-267.5)<0.001
      Sampling gestation 14+0-17+6 weeks (Mean, SD)112 (
      • Hammad S.M.
      • Al Gadban M.M.
      • Semler A.J.
      • Klein R.L.
      Sphingosine 1-phosphate distribution in human plasma: associations with lipid profiles.
      )
      114 (
      • Rana S.
      • Lemoine E.
      • Granger J.P.
      • Karumanchi S.A.
      Preeclampsia: Pathophysiology, Challenges, and Perspectives.
      )
      114 (
      • Rana S.
      • Lemoine E.
      • Granger J.P.
      • Karumanchi S.A.
      Preeclampsia: Pathophysiology, Challenges, and Perspectives.
      )
      0.51
      Sampling gestation 18+0-24+6 weeks (Mean, SD)156 (
      • Melland-Smith M.
      • Ermini L.
      • Chauvin S.
      • Craig-Barnes H.
      • Tagliaferro A.
      • Todros T.
      • et al.
      Disruption of sphingolipid metabolism augments ceramide-induced autophagy in preeclampsia.
      )
      153 (
      • Hsu S.C.
      • Chang J.H.
      • Hsu Y.P.
      • Bai K.J.
      • Huang S.K.
      • Hsu C.W.
      Circulating sphingosine-1-phosphate as a prognostic biomarker for community-acquired pneumonia.
      )
      154 (
      • Obinata H.
      • Hla T.
      Sphingosine 1-phosphate and inflammation.
      )
      0.73
      Birthweight (mean, SD)3189, 339.11779.7, 687.22991.1, 347.5<0.001
      Parity (Median, IQR)1, 21, 11, 10.56
      Fetal sex male22 (42)9 (56)12 (43)0.049
      BMI (Median, IQR)25.1, 3.425.2, 3.825.7, 5.30.49
      Chronic Hypertension47 (90)22 (81)13 (81)0.447
      Mean plasma S1P concentrations were similar to previously reported ranges in pregnant women from LC-MS experiments (0.339 (CI 0.308-0.37) nmol/ml) (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ). Mean plasma sphingosine was also similar (6.77 (CI 4.94- 8.61) nmol/l) (
      • Spijkers L.J.
      • van den Akker R.F.
      • Janssen B.J.
      • Debets J.J.
      • De Mey J.G.
      • Stroes E.S.
      • et al.
      Hypertension is associated with marked alterations in sphingolipid biology: a potential role for ceramide.
      ). S1P and sphingosine levels were not significantly affected by gestational age at sampling and were not significantly different between sampling points in women with longitudinal measurements (supp Fig. 4). There were no differences between plasma S1P or sphingosine concentrations between uncomplicated outcome, preterm preeclampsia, or term preeclampsia groups at either 14+0-17+6 weeks or 18+0-24+6 weeks (Fig. 1); this did not alter with inclusion of longitudinal measurements from the same woman. There were some unexplained outliers in the dataset with increased S1P or sphingosine compared to the population average. Removal of outliers outside 95% confidence interval did not affect the absence of relationship between uncomplicated and women with preeclampsia outcome.
      Figure thumbnail gr1
      Figure 1S1P(A) and sphingosine(B) concentrations at 2 different time points in women with Uncomplicated outcome, Preterm preeclampsia, Term preeclampsia. S1P 14+0-17+6 n=61, 18+0-24+6 n=67; Sphingosine 14=0-17+6 n=61 18+0-24+6 n=65.
      We examined the relationship between S1P and sphingosine to determine if outliers were the result of correlate S1P and sphingosine (supp Fig. 5). One value of sphingosine (35.66ng/ml) was >10x more than all other sphingosine values and was unrelated to S1P concentrations, as a result it was excluded from further analysis. Using regression, we tested the relationship between S1P and sphingosine which was significant. For every 0.1 ng/ml increase in sphingosine, S1P increased by 18.08ng/ml (95%CI 9.73-26.44, p<0.0001). However, this relationship was unaffected by different pregnancy outcomes (interaction term p>0.05) (Fig. 2).
      Figure thumbnail gr2
      Figure 2Relationship between Sphingosine and S1P in pregnancies with uncomplicated pregnancies (blue dots), preterm preeclampsia (red dots) and term preeclampsia (purple dots). Lines demonstrate fitted values and are in same colour for each outcome; only earliest gestation sample for each participant included. Relationship between Sphingosine and S1P not significantly different between pregnancy outcome groups (interaction term p>0.05).
      There was no group difference between the correlation of S1P with PlGF at either 14+0-17+6 weeks or 18+)-24+6 weeks (t test logged values; p=0.727 and p=0.789). There was no relationship between sphingosine with PlGF at either 14-17+6 weeks or 18-24 weeks (t test logged values; p=0.513 and 0.527). There was no significant association between S1P or sphingosine and PlGF (which is associated with later development of preeclampsia) at either 14-18 weeks or 18-24 weeks (Fig. 3).
      Figure thumbnail gr3
      Figure 3Relationship between Log PlGF and S1P (A/B) and sphingosine (C/D) concentrations. PlGF concentrations <5th centile (red dots) and ≥5th centile (blue dots). Panels A&C include measurements taken between 14+0-17+6 weeks and panels B&D include measurements between 18+0-24+6 weeks; for women with longitudinal samples only the earliest sample is included.

      Discussion

      Our study has demonstrated that in a cohort of pregnant women who are at increased risk of developing preeclampsia, S1P and sphingosine levels do not appear to be different before the development of disease.
      Mean plasma S1P levels for the whole cohort were lower than in previously reported non-pregnant populations (reported mean plasma levels 0.75 ± 0.16 nmol/mL (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ) to 1nmol/ml , but similar to other published studies in pregnant women (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ,
      • Melland-Smith M.
      • Ermini L.
      • Chauvin S.
      • Craig-Barnes H.
      • Tagliaferro A.
      • Todros T.
      • et al.
      Disruption of sphingolipid metabolism augments ceramide-induced autophagy in preeclampsia.
      ,
      • Charkiewicz K.
      • Goscik J.
      • Blachnio-Zabielska A.
      • Raba G.
      • Sakowicz A.
      • Kalinka J.
      • et al.
      Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
      ). Lower plasma S1P concentrations contrasts with pregnancy effects on cholesterol, triglycerides, LDLs and HDL which all increase in pregnancy (
      • Piechota W.
      • Staszewski A.
      Reference ranges of lipids and apolipoproteins in pregnancy.
      ,
      • Wiznitzer A.
      • Mayer A.
      • Novack V.
      • Sheiner E.
      • Gilutz H.
      • Malhotra A.
      • et al.
      Association of lipid levels during gestation with preeclampsia and gestational diabetes mellitus: a population-based study.
      ). Age, BMI and smoking have not been shown to significantly effect S1P concentrations (
      • Moritz E.
      • Wegner D.
      • Gross S.
      • Bahls M.
      • Dorr M.
      • Felix S.B.
      • et al.
      Reference intervals for serum sphingosine-1-phosphate in the population-based Study of Health in Pomerania.
      ) so the younger age and increased BMI of the cohort compared to published non-pregnant cohorts is unlikely to be responsible for the lower concentrations. Differences in sample handling between centres have been reported to be responsible for some difference in reported concentrations (
      • Daum G.
      • Winkler M.
      • Moritz E.
      • Muller T.
      • Geffken M.
      • von Lucadou M.
      • et al.
      Determinants of Serum- and Plasma Sphingosine-1-Phosphate Concentrations in a Healthy Study Group.
      ), but this would seem unlikely here as our findings and other pregnancy cohorts from geographically distant investigators are similar. Pregnant women have a lower haematocrit than the general population and S1P has been shown to correlate with haematocrit (
      • Daum G.
      • Winkler M.
      • Moritz E.
      • Muller T.
      • Geffken M.
      • von Lucadou M.
      • et al.
      Determinants of Serum- and Plasma Sphingosine-1-Phosphate Concentrations in a Healthy Study Group.
      ), but the strength of relationship is insufficient to explain the significant differences observed and at present it remains unclear why plasma concentrations of S1P in pregnancy seem consistently lower than the non-pregnant population. . SIP effects are partially dependent on carrier binding with potentially longer lasting effects when bound to HDL rather than albumin (
      • Wilkerson B.A.
      • Grass G.D.
      • Wing S.B.
      • Argraves W.S.
      • Argraves K.M.
      Sphingosine 1-phosphate (S1P) carrier-dependent regulation of endothelial barrier: high density lipoprotein (HDL)-S1P prolongs endothelial barrier enhancement as compared with albumin-S1P via effects on levels, trafficking, and signaling of S1P1.
      ), so it should be noted that different S1P concentrations alone do not fully explain its circulating biological effects (
      • Fakhr Y.
      • Brindley D.N.
      • Hemmings D.G.
      Physiological and pathological functions of sphingolipids in pregnancy.
      ).
      Previous studies examining the relationship between plasma S1P and preeclampsia have reported inconsistent findings. In a similar longitudinal study Dobierzewska et al (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ) examined S1P levels in first, second and third trimester women, but failed to demonstrate any significant differences between women with uncomplicated outcomes or those who developed preeclampsia at any gestation which is consistent with our findings. This contrasted with findings from Melland et al. (
      • Melland-Smith M.
      • Ermini L.
      • Chauvin S.
      • Craig-Barnes H.
      • Tagliaferro A.
      • Todros T.
      • et al.
      Disruption of sphingolipid metabolism augments ceramide-induced autophagy in preeclampsia.
      ) who found lower S1P concentrations (preeclampsia 0.263nmol/ml versus control 0.461nmol/ml). However, this was a small study of only 10 patients and analysis was performed in serum not plasma in which concentrations are higher (
      • Moritz E.
      • Wegner D.
      • Gross S.
      • Bahls M.
      • Dorr M.
      • Felix S.B.
      • et al.
      Reference intervals for serum sphingosine-1-phosphate in the population-based Study of Health in Pomerania.
      ). Charkiewicz et al. also demonstrated that in women who had been diagnosed with preeclampsia, S1P levels were significantly higher when compared to normal pregnant women (
      • Charkiewicz K.
      • Goscik J.
      • Blachnio-Zabielska A.
      • Raba G.
      • Sakowicz A.
      • Kalinka J.
      • et al.
      Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
      ). Similarly, this was a relatively small study, but S1P concentrations (0.177nmol/ml +/-0.014) were also significantly lower than all the other reported pregnancy concentrations. There are no apparent differences in the normal pregnancies between the cohorts, so the levels reported may reflect differences in measurement methodology which has been suggested to be the cause of differences in measured concentrations between different non-pregnancy cohorts (
      • Moritz E.
      • Wegner D.
      • Gross S.
      • Bahls M.
      • Dorr M.
      • Felix S.B.
      • et al.
      Reference intervals for serum sphingosine-1-phosphate in the population-based Study of Health in Pomerania.
      ),(
      • Fakhr Y.
      • Brindley D.N.
      • Hemmings D.G.
      Physiological and pathological functions of sphingolipids in pregnancy.
      ). It is also possible that our lower concentrations and those of Dobierzewska et al (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ) are the result of our patient populations pre-existing risk status being different from other published studies that report a rise in S1P. Melland et al. (
      • Melland-Smith M.
      • Ermini L.
      • Chauvin S.
      • Craig-Barnes H.
      • Tagliaferro A.
      • Todros T.
      • et al.
      Disruption of sphingolipid metabolism augments ceramide-induced autophagy in preeclampsia.
      ) and Dobierzewska et al (
      • Dobierzewska A.
      • Soman S.
      • Illanes S.E.
      • Morris A.J.
      Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
      ) and do not include information on prior pregnancy risk status; Charkiewicz et al. (
      • Charkiewicz K.
      • Goscik J.
      • Blachnio-Zabielska A.
      • Raba G.
      • Sakowicz A.
      • Kalinka J.
      • et al.
      Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
      ) excluded women with previous chronic hypertension at time of diagnosis of preeclampsia which clearly differs from our cohort.
      As with S1P, mean plasma sphingosine plasma concentrations were marginally lower than the concentrations in published non-pregnant population (
      • Hammad S.M.
      • Al Gadban M.M.
      • Semler A.J.
      • Klein R.L.
      Sphingosine 1-phosphate distribution in human plasma: associations with lipid profiles.
      ),(
      • Savica R.
      • Murray M.E.
      • Persson X.M.
      • Kantarci K.
      • Parisi J.E.
      • Dickson D.W.
      • et al.
      Plasma sphingolipid changes with autopsy-confirmed Lewy Body or Alzheimer's pathology.
      ), but significantly lower that the levels found by Charkiewicz et al. (
      • Charkiewicz K.
      • Goscik J.
      • Blachnio-Zabielska A.
      • Raba G.
      • Sakowicz A.
      • Kalinka J.
      • et al.
      Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
      ) in normal and preeclampsia pregnancies. It should be noted that the Charkiewicz measurements were from women who had developed clinical preeclampsia, however this does not explain the very high levels seen in normal outcome pregnancies in that study which is not in keeping with published ranges of sphingosine fractions in plasma (
      • Hammad S.M.
      • Al Gadban M.M.
      • Semler A.J.
      • Klein R.L.
      Sphingosine 1-phosphate distribution in human plasma: associations with lipid profiles.
      ). It is possible that due to the magnitude of difference, reported concentrations are the result of different measurement techniques, but as with S1P we are unable to fully explain these different results.
      In our data we noticed significantly increased levels of S1P and sphingosine in some subjects and on interrogating the data there was a correlation between the two sphingolipids. To our knowledge this continuous relationship is a novel finding that has not been previously observed in plasma in any patient group. As with individual S1P and sphingosine measurements, there was no relationship with combined measurements and pregnancy outcome.

      S1P and sphingosine as predictors of preeclampsia

      Our study’s main strengths are the cohort size and the inclusion of women who had uncomplicated outcomes, preterm preeclampsia or term preeclampsia which allowed us to test the potential use of S1P and sphingosine as early biomarkers of different phenotypes the disease. Although our cohort is large a limitation of our study is that it did not include any women who were low risk at the outset of pregnancy, and it is therefore possible that the relationship between S1P and sphingosine and preeclampsia differ in our higher risk cohort. A further study limitation arises from the fact that the samples were taken early in pregnancy before development of the disease features did not allow for near disease measurement and therefore it remains uncertain if these sphingolipids could be useful when combined with existing biomarkers of preeclampsia, such as PlGF. PlGF is a member of the VEGF family and, when maternal plasma concentrations are low, is associated with increased chances of preeclampsia/fetal growth restriction (
      • Duhig K.E.
      • Myers J.
      • Seed P.T.
      • Sparkes J.
      • Lowe J.
      • Hunter R.M.
      • et al.
      Placental growth factor testing to assess women with suspected pre-eclampsia: a multicentre, pragmatic, stepped-wedge cluster-randomised controlled trial.
      ). In later pregnancy low concentrations are predictive of imminent disease , but very low levels between 14 and 24 weeks are known to be associated with an increased risk of preterm preeclampsia (
      • Myers J.E.
      • Kenny L.C.
      • McCowan L.M.
      • Chan E.H.
      • Dekker G.A.
      • Poston L.
      • et al.
      Angiogenic factors combined with clinical risk factors to predict preterm pre-eclampsia in nulliparous women: a predictive test accuracy study.
      ). We therefore examined the relationship between S1P and sphingosine and PlGF and were unable to demonstrate any overlap either continuously or using very low levels of PlGF as indicative of potentially high-risk pregnancies.

      Conclusion

      Plasma S1P concentrations are lower in pregnancy, but previously reported changes in S1P concentrations prior to the development of preeclampsia were not present in our well characterised cohort of term and preterm preeclampsia. Circulating S1P or sphingosine concentrations in early pregnancy are not associated with later onset of preeclampsia in women at high pre-pregnancy risk of developing the disease.

      Data availability

      Due to the identifiable nature of human data within this manuscript it is not available for secondardy analyses.

      Acknowledgements

      Women attending MAViS who donated samples.
      Riainne Wallworth, technician.
      Neil O’Hara for excellent technical support.

      References

        • Huang W.C.
        • Liang J.
        • Nagahashi M.
        • Avni D.
        • Yamada A.
        • Maceyka M.
        • et al.
        Sphingosine-1-phosphate phosphatase 2 promotes disruption of mucosal integrity, and contributes to ulcerative colitis in mice and humans.
        FASEB J. 2016; 30: 2945-2958
        • Pyne S.
        • Lee S.C.
        • Long J.
        • Pyne N.J.
        Role of sphingosine kinases and lipid phosphate phosphatases in regulating spatial sphingosine 1-phosphate signalling in health and disease.
        Cell Signal. 2009; 21: 14-21
        • Newton J.
        • Lima S.
        • Maceyka M.
        • Spiegel S.
        Revisiting the sphingolipid rheostat: Evolving concepts in cancer therapy.
        Exp Cell Res. 2015; 333: 195-200
        • Westwood M.
        • Al-Saghir K.
        • Finn-Sell S.
        • Tan C.
        • Cowley E.
        • Berneau S.
        • et al.
        Vitamin D attenuates sphingosine-1-phosphate (S1P)-mediated inhibition of extravillous trophoblast migration.
        Placenta. 2017; 60: 1-8
        • Mizugishi K.
        • Li C.
        • Olivera A.
        • Bielawski J.
        • Bielawska A.
        • Deng C.X.
        • et al.
        Maternal disturbance in activated sphingolipid metabolism causes pregnancy loss in mice.
        J Clin Invest. 2007; 117: 2993-3006
        • Nagahashi M.
        • Abe M.
        • Sakimura K.
        • Takabe K.
        • Wakai T.
        The role of sphingosine-1-phosphate in inflammation and cancer progression.
        Cancer Sci. 2018; 109: 3671-3678
        • Rana S.
        • Lemoine E.
        • Granger J.P.
        • Karumanchi S.A.
        Preeclampsia: Pathophysiology, Challenges, and Perspectives.
        Circ Res. 2019; 124: 1094-1112
        • Hammad S.M.
        • Al Gadban M.M.
        • Semler A.J.
        • Klein R.L.
        Sphingosine 1-phosphate distribution in human plasma: associations with lipid profiles.
        J Lipids. 2012; 2012180705
        • Dobierzewska A.
        • Soman S.
        • Illanes S.E.
        • Morris A.J.
        Plasma cross-gestational sphingolipidomic analyses reveal potential first trimester biomarkers of preeclampsia.
        PLoS One. 2017; 12e0175118
        • Halilbasic E.
        • Fuerst E.
        • Heiden D.
        • Japtok L.
        • Diesner S.C.
        • Trauner M.
        • et al.
        Plasma Levels of the Bioactive Sphingolipid Metabolite S1P in Adult Cystic Fibrosis Patients: Potential Target for Immunonutrition?.
        Nutrients. 2020; 12
        • Hsu S.C.
        • Chang J.H.
        • Hsu Y.P.
        • Bai K.J.
        • Huang S.K.
        • Hsu C.W.
        Circulating sphingosine-1-phosphate as a prognostic biomarker for community-acquired pneumonia.
        PLoS One. 2019; 14e0216963
        • Obinata H.
        • Hla T.
        Sphingosine 1-phosphate and inflammation.
        Int Immunol. 2019; 31: 617-625
        • Melland-Smith M.
        • Ermini L.
        • Chauvin S.
        • Craig-Barnes H.
        • Tagliaferro A.
        • Todros T.
        • et al.
        Disruption of sphingolipid metabolism augments ceramide-induced autophagy in preeclampsia.
        Autophagy. 2015; 11: 653-669
        • Charkiewicz K.
        • Goscik J.
        • Blachnio-Zabielska A.
        • Raba G.
        • Sakowicz A.
        • Kalinka J.
        • et al.
        Sphingolipids as a new factor in the pathomechanism of preeclampsia - Mass spectrometry analysis.
        PLoS One. 2017; 12e0177601
        • Shawkat E.
        • Mistry H.
        • Chmiel C.
        • Webster L.
        • Chappell L.
        • Johnstone E.D.
        • et al.
        The effect of labetalol and nifedipine MR on blood pressure in women with chronic hypertension in pregnancy.
        Pregnancy Hypertens. 2018; 11: 92-98
        • Brown M.A.
        • Magee L.A.
        • Kenny L.C.
        • Karumanchi S.A.
        • McCarthy F.P.
        • Saito S.
        • et al.
        The hypertensive disorders of pregnancy: ISSHP classification, diagnosis & management recommendations for international practice.
        Pregnancy Hypertens. 2018; 13: 291-310
        • Kendall A.C.
        • Pilkington S.M.
        • Massey K.A.
        • Sassano G.
        • Rhodes L.E.
        • Nicolaou A.
        Distribution of bioactive lipid mediators in human skin.
        J Invest Dermatol. 2015; 135: 1510-1520
        • Bielawski J.
        • Szulc Z.M.
        • Hannun Y.A.
        • Bielawska A.
        Simultaneous quantitative analysis of bioactive sphingolipids by high-performance liquid chromatography-tandem mass spectrometry.
        Methods. 2006; 39: 82-91
        • Kendall A.C.
        • Kiezel-Tsugunova M.
        • Brownbridge L.C.
        • Harwood J.L.
        • Nicolaou A.
        Lipid functions in skin: Differential effects of n-3 polyunsaturated fatty acids on cutaneous ceramides, in a human skin organ culture model.
        Biochim Biophys Acta Biomembr. 2017; 1859: 1679-1689
        • Verlohren S.
        • Herraiz I.
        • Lapaire O.
        • Schlembach D.
        • Zeisler H.
        • Calda P.
        • et al.
        New gestational phase-specific cutoff values for the use of the soluble fms-like tyrosine kinase-1/placental growth factor ratio as a diagnostic test for preeclampsia.
        Hypertension. 2014; 63: 346-352
        • Spijkers L.J.
        • van den Akker R.F.
        • Janssen B.J.
        • Debets J.J.
        • De Mey J.G.
        • Stroes E.S.
        • et al.
        Hypertension is associated with marked alterations in sphingolipid biology: a potential role for ceramide.
        PLoS One. 2011; 6e21817
        • Piechota W.
        • Staszewski A.
        Reference ranges of lipids and apolipoproteins in pregnancy.
        Eur J Obstet Gynecol Reprod Biol. 1992; 45: 27-35
        • Wiznitzer A.
        • Mayer A.
        • Novack V.
        • Sheiner E.
        • Gilutz H.
        • Malhotra A.
        • et al.
        Association of lipid levels during gestation with preeclampsia and gestational diabetes mellitus: a population-based study.
        Am J Obstet Gynecol. 2009; 201: 482 e1-8
        • Moritz E.
        • Wegner D.
        • Gross S.
        • Bahls M.
        • Dorr M.
        • Felix S.B.
        • et al.
        Reference intervals for serum sphingosine-1-phosphate in the population-based Study of Health in Pomerania.
        Clin Chim Acta. 2017; 468: 25-31
        • Daum G.
        • Winkler M.
        • Moritz E.
        • Muller T.
        • Geffken M.
        • von Lucadou M.
        • et al.
        Determinants of Serum- and Plasma Sphingosine-1-Phosphate Concentrations in a Healthy Study Group.
        TH Open. 2020; 4: e12-e19
        • Wilkerson B.A.
        • Grass G.D.
        • Wing S.B.
        • Argraves W.S.
        • Argraves K.M.
        Sphingosine 1-phosphate (S1P) carrier-dependent regulation of endothelial barrier: high density lipoprotein (HDL)-S1P prolongs endothelial barrier enhancement as compared with albumin-S1P via effects on levels, trafficking, and signaling of S1P1.
        J Biol Chem. 2012; 287: 44645-44653
        • Fakhr Y.
        • Brindley D.N.
        • Hemmings D.G.
        Physiological and pathological functions of sphingolipids in pregnancy.
        Cell Signal. 2021; 85110041
        • Savica R.
        • Murray M.E.
        • Persson X.M.
        • Kantarci K.
        • Parisi J.E.
        • Dickson D.W.
        • et al.
        Plasma sphingolipid changes with autopsy-confirmed Lewy Body or Alzheimer's pathology.
        Alzheimers Dement (Amst). 2016; 3: 43-50
        • Duhig K.E.
        • Myers J.
        • Seed P.T.
        • Sparkes J.
        • Lowe J.
        • Hunter R.M.
        • et al.
        Placental growth factor testing to assess women with suspected pre-eclampsia: a multicentre, pragmatic, stepped-wedge cluster-randomised controlled trial.
        Lancet. 2019; 393: 1807-1818
        • Myers J.E.
        • Kenny L.C.
        • McCowan L.M.
        • Chan E.H.
        • Dekker G.A.
        • Poston L.
        • et al.
        Angiogenic factors combined with clinical risk factors to predict preterm pre-eclampsia in nulliparous women: a predictive test accuracy study.
        BJOG. 2013; 120: 1215-1223