Untargeted analysis of plasma samples from pre-eclamptic women reveals polar and apolar changes in the metabolome

Abstract

Introduction: Pre-eclampsia is a hypertensive gestational disorder that affects approximately 5% of all pregnancies.

Objectives: As the pathophysiological processes of pre-eclampsia are still uncertain, the present case-control study explored underlying metabolic processes characterising this disease.

Methods: Maternal peripheral plasma samples were collected from pre-eclamptic (n = 32) and healthy pregnant women (n = 35) in the third trimester. After extraction, high-resolution mass spectrometry-based untargeted metabolomics was used to profile polar and apolar metabolites and the resulting data were analysed via uni- and multivariate statistical approaches.

Results: The study demonstrated that the metabolome undergoes substantial changes in pre-eclamptic women. Amongst the most discriminative metabolites were hydroxyhexacosanoic acid, diacylglycerols, glycerophosphoinositols, nicotinamide adenine dinucleotide metabolites, bile acids and products of amino acid metabolism.

Conclusions: The putatively identified compounds provide sources for novel hypotheses to help understanding of the underlying biochemical pathology of pre-eclampsia.

Keywords: Metabolic profiling; Metabolomics; Pre-eclampsia; Pregnancy, human.

Fig. 1

Multivariate analysis based on all detected ions: a OPLS-DA score plot of control (no hypertensive disease, nHD, n = 35) and pre-eclampsia (PET, n = 32) samples for all variables. b OPLS-DA score plot of training and prediction set for control (no hypertensive disease, nHD) and pre-eclampsia (PET) samples for all variables. c Receiver operator characteristic (ROC) curves for all variables. The figure shows the true positive fraction (TPF) with upper and lower 95% confidence intervals. The AUC is 0.922 with a standard error of 0.06

Fig. 2

Multivariate analysis based on the 35 most predictive ions: a OPLS-DA score plot of control (no hypertensive disease, nHD, n = 35) and pre-eclampsia (PET, n = 32) samples for the 35 most predictive ions. b OPLS-DA score plots of training and prediction set for control (no hypertensive disease, nHD) and pre-eclampsia (PET) samples for the 35 most predictive ions. c Receiver operator characteristic (ROC) curves for the 35 most predictive ions. The figure shows the true positive fraction (TPF) with upper and lower 95% confidence intervals. The AUC is 0.964 with a standard error of 0.04

Fig. 3

Scatter dot plots of potential biomarkers in the polar (a) and apolar (b) phase. All shown biomarkers fulfil criteria of both MVA and UVA. MVA criteria were a VIP value > 1 in the OPLS-DA model shown in Fig. 1c. Criteria for UVA were a q-value < 0.05 (*), a fold change > 1.5 and CV(QC) < 30%. Dots show log(1 + normalised intensity) of control (n = 35) and pre-eclampsia (PET, n = 32) samples. Error bars represent median and interquartile ranges. For more details see supplemental data (Tables 1, 2 and Table 3)