Plasma metabolites are altered before and after diagnosis of preeclampsia or fetal growth restriction

Abstract

Metabolomics is the study of small molecules (metabolites), within cells, tissues and biofluids. Maternal metabolites can provide important insight into the health and development of both mother and fetus throughout pregnancy. This study assessed metabolic profiles in the maternal circulation prior to and at the time of diagnosis of preeclampsia and fetal growth restriction. Maternal plasma samples were collected from two independent cohorts: (1) Established disease cohort: 50 participants diagnosed with early-onset preeclampsia (< 34 weeks' gestation), 14 with early-onset fetal growth restriction, and 25 gestation-matched controls. (2) Prospective cohort, collected at 36 weeks' gestation before diagnosis: 17 participants later developed preeclampsia, 49 delivered infants with fetal growth restriction (birthweight < 5th centile), and 72 randomly selected controls. Metabolic evaluation was performed by Metabolomics Australia on the Agilent 6545 QTOF Mass Spectrometer. In the established disease cohort, 77 metabolites were altered in circulation from participants with preeclampsia - increased L-cysteine (3.73-fold), L-cystine (3.28-fold), L-acetylcarnitine (2.57-fold), and carnitine (1.53-fold) (p < 0.05). There were 53 metabolites dysregulated in participants who delivered a fetal growth restriction infant-including increased levulinic acid, citric acid (1.93-fold), and creatine (1.14-fold) (p < 0.05). In the prospective cohort, 30 metabolites were altered in participants who later developed preeclampsia at term - reduced glutaric acid (0.85-fold), porphobilinogen (0.77-fold) and amininohippuric acid (0.82-fold) (p < 0.05) was observed. There were 5 metabolites altered in participants who later delivered a fetal growth restriction infant - including reduced 3-methoxybenzenepropanoic acid (p < 0.05). Downstream pathway analysis revealed aminoacyl-tRNA biosynthesis to be most significantly altered in the established cohort in preeclampsia (13/48 hits, p < 0.001) and fetal growth restriction (7/48 hits, p < 0.001). The predictive cohort showed no significant pathway alterations. This study observed altered metabolites in maternal plasma collected before and after diagnosis of a preeclampsia or fetal growth restriction. While a significant number of metabolites were altered with established disease, few changes were observed in the predictive cohort. Thus, metabolites measured in this study may not be useful as predictors of preeclampsia or fetal growth restriction.

Figure 1

Hierarchical cluster analysis via heatmaps and volcano plots revealed metabolite profile differences in the metabolome of plasma collected at < 34 weeks’ gestation from pregnancies complicated by early-onset preeclampsia (A and C) and fetal growth restriction (B and D). (A) Heatmap and dendrogram using Euclidean distance measurement and ward clustering method of top 25 significant metabolites for control (Blue), and preeclampsia (Red) pregnancies. (B) Heatmap and dendrogram using Euclidean distance measurement and ward clustering method of top 25 significant metabolites for control (Blue), and preeclampsia (Red) samples collected from Established disease cohort. Colours for samples are denoted in the top x-axis of both (A) and (B). All metabolites have been scaled. Heatmap scales are denoted by red being higher, blue being lower, with white residing in the middle; this scale is used for each of the metabolites listed on the y-axis. Ward clustering algorithm used for dendrograms. (C) Established fetal growth restriction volcano plot, x-axis indicates the fold change (FC) in Log2 format, whilst the y-axis is t-test significance in − Log10. (D) Established fetal growth restriction volcano plot, x-axis indicates the fold change (FC) in Log2 format, whilst the y-axis is t-test significance in − Log10. In both instances of (C) and (D) standard t-test was conducted for p value calculations, significance was considered at p value of 0.1 along with a FC of 2.0. Ratios of Fold changes are calculated as disease/ control with blue indicating decreased fold change and red indicating an increased fold change.

Figure 2

Heatmaps and volcano plots of metabolite data collected from the Biomarker and Ultrasound Measures for Preventable Stillbirth (BUMPS). Heatmaps, dendrograms and volcano plots of maternal plasma collected at 36 weeks’ gestation prior to diagnosis of term preeclampsia (A and C) or delivered a fetal growth restricted infant (B and D), compared to controls. (A) Heatmap and dendrogram using Euclidean distance measurement and ward clustering method of top 25 significant Metabolites for control (Blue), and preeclampsia (Red) pregnancies. (B) Heatmap and dendrogram using Euclidean distance measurement and ward clustering method of top 25 significant Metabolites for control (Blue), and preeclampsia (Red) samples collected from BUMPs cohort. Colours for samples are denoted in the top x-axis of both (A) and (B). All metabolites have been scaled. Heatmap scales are denoted by red being higher, blue being lower, with white residing in the middle; this scale is used for each of the metabolites listed on the y-axis. Ward clustering algorithm used for dendrograms. (C) Established fetal growth restriction volcano plot, x-axis indicates the fold change (FC) in Log2 format, whilst the y-axis is t-test significance in − Log10. (D) BUMPS fetal growth restriction volcano plot, x-axis indicates the fold change (FC) in Log2 format, whilst the y-axis is t-test significance in − Log10. In both instances of (C) and (D) standard t-test was conducted for p value calculations, significance was considered at p value of 0.1 along with a FC of 2.0. Ratios of Fold changes are calculated as disease/ control with blue indicating decreased fold change and red indicating an increased fold change.

Figure 3

Venn diagram of the metabolite cross over between established preeclampsia (Blue), Established fetal growth restriction (Yellow), BUMPS preeclampsia (Purple), and BUMPS fetal growth restriction (orange). Numbers indicate the number of significantly different metabolites that are exclusive to that cohort and outcome. Where circles overlap, indicates the number of metabolites that are significantly different from controls that are shared between outcomes and/or cohort.