Gestational diabetes mellitus, intrauterine growth restriction, and preeclamptic toxemia are common pregnancy complications that can have detrimental effects on morbidity and mortality of the mother and fetus as well as long-term health outcomes. Although they are distinct conditions, they may occur together and are often considered together as they share a common etiology of inadequate placental perfusion. The discovery and study of preventative treatments is hampered by a lack of effective screening tools to accurately identify women at the highest risk of disease. Metabolomics, an omic science, is the global quantitative assessment of endogenous metabolites within a biological system. It has proven to be a rapid approach in the identification of biomarkers predictive of the outcome of a pathological condition and the individual’s response to a pharmacological treatment. We review the current and potential applications of metabolomics in maternal–fetal medicine, focusing on its use as a biomarker for great obstetrical syndromes diagnosis.
Introduction
Common complications of pregnancy are associated with placental dysfunction and include gestational diabetes mellitus (GDM), intrauterine growth restriction (IUGR), and preeclamptic toxemia (PET). In these pathologies, the metabolic and blood vessel adaptation to pregnancy is not optimal, leading to hypertension, poor placental blood perfusion, and failure to respond properly to hormones such as insulin. These effects can not only have immediate consequences for the health of the mother and baby but also may impact on their future health. From the epidemiological point of view, it has been reported that 10–20% of pregnancies are complicated by some form of hypertension in the advanced economies countries . Unfortunately, often no adequate predictive methods are available for these disorders, thus posing a clear need to develop noninvasive and objective methods in prenatal diagnosis and monitoring.
Metabolomics is a relatively new research field based on the systematic study of the complete set of metabolites (metabolome) in a biological sample . In particular, metabolomics is focused on the detection, identification, and quantification of hundreds/thousands of low-molecular-weight metabolites, intermediate or products of metabolism, in cells, tissues, or biofluids. Biological samples have unique and characteristic biochemical compositions that change in response to physiological or pathophysiological stimuli to generate a metabolic “fingerprint.” Thus, the metabolome snapshot produced by metabolomics reflects the metabolic responses of living systems to disease , toxicological or nutritional stimuli , thus facilitating the understanding of the mechanisms of biological and biochemical processes in complex systems.
The metabolomics approach can provide also an important diagnostic aid to diseases. In this particular case, the fundamental objective is to identify and validate biomarkers that play important roles in estimating the risk of different pathologies, in prognosis, monitoring, and assessing therapeutic responses. The approaches used toward this aim include a combined used of spectrometric and spectroscopic techniques and computer programs. The techniques commonly employed are nuclear magnetic resonance (NMR) spectroscopy, gas or liquid chromatography–mass spectrometry (MS), Fourier transform infrared spectrometry, and capillary electrophoresis–MS. One of the difficulties with all these techniques is that they provide complex datasets, due to the large numbers of metabolites generated from multiple subjects. Thus, the analysis and interpretation of data rely on multivariate methods to produce classification models for testing the predictive ability of identified biomarkers .
Metabolomics appears to be a promising technique also in pregnancy. Several metabolic profiling studies have been carried out mostly based on amniotic fluid, blood plasma, and urines, pointing out the potential of biofluids metabolic profiling for prenatal biomarker identification . This review focuses on the applications of this technology in the context of prenatal research, attempting to meet specific medical needs of diagnosis of the great obstetrical syndromes ( Table 1 ).
| Obstetrical syndromes | References | Sample | Metabolomic analysis | Key metabolites affected |
|---|---|---|---|---|
| GDM | Graca et al., 2010 | Amniotic fluid | 1 H-NMR | Glucose, acetate, creatinine, formate, glutamate, glycine, glycerophosphocholine, proline, serine, and taurine |
| Diaz et al., 2011 | Urine Plasma | 1 H-NMR | 3-hydroxyisovalerate and 2-hydroxyisobutyrate in urine Trimethylamine N-oxide and betaine in plasma | |
| Graca et al., 2012 | Urine and amniotic fluid | UPLC-MS | No predictive metabolites were detected | |
| Sachse et al., 2012 | Urine | 1 H-NMR | Lactose and citrate | |
| Logan et al., 2012 | Urine | 1 H-NMR | Glucose, formate, fumarate, succinate, and citrate | |
| Dani et al., 2013 | Cord serum | 1 H-NMR | Glucose, pyruvate, histidine, alanine, valine, methionine, arginine, lysine, hypoxanthine, lipoproteins, and lipids | |
| PET | Kenny et al., 2005 | Plasma | GC-tof-MS | 3 unidentified biomarkers |
| Kenny et al., 2008 | Plasma | UPLC–LTQ–Orbitrap–MS | Uric acid, 2-oxoglutarate, glutamate and alanine | |
| Dunn et al., 2009 | Placental villous fragments | UPLC–LTQ–Orbitrap–MS GC–ToF–MS | Glutamate, glutamine, tryptophan, leukotriene, and prostaglandin | |
| Kenny et al., 2010 | Plasma | UPLC-MS | 14 metabolites, belonging to the class of amino acids, carbohydrates, fatty acids, keto or hydroxy acids, lipids, phospholipids and steroids | |
| Odibo et al., 2011 | Blood | LC-MS/MS | Hydroxylate carnitine esters, alanine, hydroxyproline, phenylalanine, arginine, and glutamate | |
| Bahado-Singh et al., 2012 | Plasma | 1 H-NMR | Citrate, glycerol, hydroxyisovalerate and methionine | |
| Bahado-Singh et al., 2013 | Serum | 1 H-NMR | Glycerol and carnitine | |
| IUGR | Nissen et al., 2011 | Plasma | 1 H-NMR | Myo-inositol and D-chiro-inositol |
| Dessì et al., 2011 | Urine | 1 H-NMR | Myo-inositol, sarcosine, carnitine and creatinine | |
| Alexandre-Gouabau et al., 2011 | Plasma | LC-MS | Proline, arginine, histidine, tyrosine and carnitine | |
| Horgan et al., 2011 | Venous cord plasma | UPLC-MS | Phenylacetylglutamine, carnitine and hydroxybutyrate | |
| Favretto et al., 2012 | Cord blood | LC-MS | Phenylalanine, tryptophan acid, and glutamate | |
| Lin et al., 2012 | Umbilical vein plasma | Q-TOF MS | Pyroglutamic acid, carnitine and creatinine | |
| van Vliet et al., 2013 | Brains | LC-QTOF-MS | Aspargine, ornithine, N-acetylaspartylglutamic acid, N-acetylaspartate, and palmitoleic acid | |
| Cosmi et al., 2013 | Umbilical vein plasma | LC – MS | Phenylalanine, sphingosine, glycerophosphocholine, valine, tryptophan, isoleucine, and proline | |
| Sanz-Cortés et al., 2013 | Umbilical vein plasma | 1 H-NMR | Glucose, acetone, glutamine, creatine, phenylalanine, tyrosine, valine, leucine, and choline |
Metabolomics and GDM
GDM is a condition that complicates about 7% of all pregnancies. It is defined as any degree of glucose intolerance with onset or first recognition during pregnancy . GDM is associated with an increased risk of complications for the mother, fetus, and newborn. Women with GDM have an increased frequency of hypertensive disorders with the need for cesarean delivery and a high risk of developing type 2 diabetes mellitus (T2DM) later in life . Risks for the fetus and newborn include macrosomia, neonatal hypoglycemia, respiratory distress syndrome, jaundice, and also long-term consequences such as T2DM, childhood obesity, and metabolic syndrome in adults . Fortunately, short-term pregnancy outcomes can be improved and the risk of later T2DM reduced through lifestyle intervention, turning the prevention of GDM into a crucial opportunity to positively impact the life and health of mother and child . In order to reduce adverse short- and long-term outcomes for mothers and offspring, research is focusing to the improvement of the identification of pregnancies at high risk for GDM and other complications.
The first study on GDM by metabolomics was published in the literature in 2010 . Here, a NMR-based metabolomics approach was employed by Graça et al. to correlate the composition of amniotic fluid (AF), collected in the second trimester of pregnancy, to the occurrence of prenatal disorders. The investigated cases were: controls, obtained from women with healthy pregnancies ( n = 82); fetal malformations (FMs, n = 27); prediagnostic GDM for women diagnosed with GDM later in their pregnancy ( n = 27); pre-preterm delivery (PTD) for women who gave birth prior to 37th gestational week ( n = 12); pre-premature rupture of the membranes for women who had prelabor rupture of membranes after 37 g.w. ( n = 34); chromosomal disorders, generally diagnosed ca. 2 weeks after amniocentesis ( n = 10). Among these disorders, FMs were found to have the highest impact on AF metabolite composition. Twenty-two metabolites were found to change significantly, confirming previous indications that malformed fetuses suffer altered energy metabolism and kidney underdevelopment. In addition, newly found changes suggested effects on protein and nucleotide sugar biosynthesis. Compared to FMs, the impact of the prediagnostic GDM condition on the composition of AF was considerably lower: a small average increase in glucose, together with slight decreases in acetate, creatinine, formate, glutamate, glycine, glycerophosphocholine, proline, serine, and taurine were observed. These results were consistent with previously reported insulin increase in AF and higher demand for protein and nucleotide biosynthesis, as well as possible changes in renal function and lipid metabolism. This study was later extended by Graça and coworkers to compare the effects of prenatal diseases on second trimester maternal urine and plasma . Following an experimental design similar to those previously adopted , the investigated cases were: controls; FMs; prediagnostic GDM; PTD; pre-premature rupture of the membranes; chromosomal disorders. The subjects who donated urine were not necessarily the same who donated blood plasma, although an overlap of 61 subjects (in a total of 198) occurred. Regarding the urine NMR metabolic profile of the prediagnostic GDM group, higher levels of 3-hydroxyisovalerate and 2-hydroxyisobutyrate levels were noted, suggesting an altered biotin status and amino acid and/or gut metabolisms. Other urinary changes suggested choline and nucleotide metabolic alterations. Regarding the plasma composition, decreases in trimethylamine oxide (TMAO) and betaine were observed.
Again, Graça et al. were the first to describe the account of untargeted ultra-performance liquid chromatography tandem mass-spectrometry (UPLC-MS) metabonomics of maternal urine and AF to probe the effects of FMs, GDM, and PTD . For FMs, considerable metabolite variations in AF and, to a lesser extent, in urine were observed, allowing the authors to advance a metabolic picture for FM pregnancies consistent with previous studies. Regarding the GDM cases, none of the models obtained for AF and maternal urine collected for the pre-diagnostic GDM group had predictive power. This result was discussed in terms of sampling characteristics (low number of cases, n = 20) and the specific experimental conditions chosen for MS methods.
1 H NMR spectroscopy was also employed by Sachse et al. to analyze the metabolic profile of urine samples collected at three time visits (gestational week 8–20; week 28 ± 2; 10–16 weeks postpartum) from women who were diagnosed with GDM. The multivariate analysis of NMR data evidenced appreciable changes in lactose and presumes hormones concentrations in urine among the visits. Differently, multivariate methods did not point out significant modifications of the pattern of urinary metabolites according to the disease. These results suggested that NMR-based metabolomics does not provide robust identification of GDM cases.
The last two studies in the literature consider the fetal metabolism of infants born by mothers with GDM, in order to evaluate the effects of maternal disease on the development of their children. Logan et al. in their study hypothesized that the metabolic profile of infants born by diabetic mothers (IDM) is different from that of infants born by healthy mothers. Therefore, the authors analyzed urine samples from 18 IDM and 12 healthy term control infants. The metabolomic analysis was performed using one-dimensional 1 H NMR spectroscopy. The statistical analysis of the data allowed to discriminate between the two groups of newborns and the most important differences with regard to the urinary metabolites were those detected for glucose, formate, fumarate, citrate, and succinate: all metabolites involved in the tricarboxylic acid (TCA) cycle. These preliminary data indicated differences in the metabolism of IDM and control infants in the newborn period.
Another similar study was done by Dani et al. who investigated the metabolome of babies born by mothers with gestational diabetes. The purpose of the study was to compare the metabolomic profile of infants of IGDMs to that of infants of healthy mothers to evaluate if differences remain despite a strict control of gestational diabetes. The metabolomic analysis was performed on serum from 30 term IGDMs and 40 controls. The results of this investigation showed that IGDMs have lower level of glucose and higher level of pyruvate, histidine, alanine, valine, methionine, arginine, lysine, hypoxanthine, lipoproteins, and lipids than controls, but showed no clinical differences. Authors concluded that prolonged fetal exposure to hyperglycemia during pregnancy can change neonatal metabolomic profile at birth without affecting the clinical course.
Metabolomics and GDM
GDM is a condition that complicates about 7% of all pregnancies. It is defined as any degree of glucose intolerance with onset or first recognition during pregnancy . GDM is associated with an increased risk of complications for the mother, fetus, and newborn. Women with GDM have an increased frequency of hypertensive disorders with the need for cesarean delivery and a high risk of developing type 2 diabetes mellitus (T2DM) later in life . Risks for the fetus and newborn include macrosomia, neonatal hypoglycemia, respiratory distress syndrome, jaundice, and also long-term consequences such as T2DM, childhood obesity, and metabolic syndrome in adults . Fortunately, short-term pregnancy outcomes can be improved and the risk of later T2DM reduced through lifestyle intervention, turning the prevention of GDM into a crucial opportunity to positively impact the life and health of mother and child . In order to reduce adverse short- and long-term outcomes for mothers and offspring, research is focusing to the improvement of the identification of pregnancies at high risk for GDM and other complications.
The first study on GDM by metabolomics was published in the literature in 2010 . Here, a NMR-based metabolomics approach was employed by Graça et al. to correlate the composition of amniotic fluid (AF), collected in the second trimester of pregnancy, to the occurrence of prenatal disorders. The investigated cases were: controls, obtained from women with healthy pregnancies ( n = 82); fetal malformations (FMs, n = 27); prediagnostic GDM for women diagnosed with GDM later in their pregnancy ( n = 27); pre-preterm delivery (PTD) for women who gave birth prior to 37th gestational week ( n = 12); pre-premature rupture of the membranes for women who had prelabor rupture of membranes after 37 g.w. ( n = 34); chromosomal disorders, generally diagnosed ca. 2 weeks after amniocentesis ( n = 10). Among these disorders, FMs were found to have the highest impact on AF metabolite composition. Twenty-two metabolites were found to change significantly, confirming previous indications that malformed fetuses suffer altered energy metabolism and kidney underdevelopment. In addition, newly found changes suggested effects on protein and nucleotide sugar biosynthesis. Compared to FMs, the impact of the prediagnostic GDM condition on the composition of AF was considerably lower: a small average increase in glucose, together with slight decreases in acetate, creatinine, formate, glutamate, glycine, glycerophosphocholine, proline, serine, and taurine were observed. These results were consistent with previously reported insulin increase in AF and higher demand for protein and nucleotide biosynthesis, as well as possible changes in renal function and lipid metabolism. This study was later extended by Graça and coworkers to compare the effects of prenatal diseases on second trimester maternal urine and plasma . Following an experimental design similar to those previously adopted , the investigated cases were: controls; FMs; prediagnostic GDM; PTD; pre-premature rupture of the membranes; chromosomal disorders. The subjects who donated urine were not necessarily the same who donated blood plasma, although an overlap of 61 subjects (in a total of 198) occurred. Regarding the urine NMR metabolic profile of the prediagnostic GDM group, higher levels of 3-hydroxyisovalerate and 2-hydroxyisobutyrate levels were noted, suggesting an altered biotin status and amino acid and/or gut metabolisms. Other urinary changes suggested choline and nucleotide metabolic alterations. Regarding the plasma composition, decreases in trimethylamine oxide (TMAO) and betaine were observed.
Again, Graça et al. were the first to describe the account of untargeted ultra-performance liquid chromatography tandem mass-spectrometry (UPLC-MS) metabonomics of maternal urine and AF to probe the effects of FMs, GDM, and PTD . For FMs, considerable metabolite variations in AF and, to a lesser extent, in urine were observed, allowing the authors to advance a metabolic picture for FM pregnancies consistent with previous studies. Regarding the GDM cases, none of the models obtained for AF and maternal urine collected for the pre-diagnostic GDM group had predictive power. This result was discussed in terms of sampling characteristics (low number of cases, n = 20) and the specific experimental conditions chosen for MS methods.
1 H NMR spectroscopy was also employed by Sachse et al. to analyze the metabolic profile of urine samples collected at three time visits (gestational week 8–20; week 28 ± 2; 10–16 weeks postpartum) from women who were diagnosed with GDM. The multivariate analysis of NMR data evidenced appreciable changes in lactose and presumes hormones concentrations in urine among the visits. Differently, multivariate methods did not point out significant modifications of the pattern of urinary metabolites according to the disease. These results suggested that NMR-based metabolomics does not provide robust identification of GDM cases.
The last two studies in the literature consider the fetal metabolism of infants born by mothers with GDM, in order to evaluate the effects of maternal disease on the development of their children. Logan et al. in their study hypothesized that the metabolic profile of infants born by diabetic mothers (IDM) is different from that of infants born by healthy mothers. Therefore, the authors analyzed urine samples from 18 IDM and 12 healthy term control infants. The metabolomic analysis was performed using one-dimensional 1 H NMR spectroscopy. The statistical analysis of the data allowed to discriminate between the two groups of newborns and the most important differences with regard to the urinary metabolites were those detected for glucose, formate, fumarate, citrate, and succinate: all metabolites involved in the tricarboxylic acid (TCA) cycle. These preliminary data indicated differences in the metabolism of IDM and control infants in the newborn period.
Another similar study was done by Dani et al. who investigated the metabolome of babies born by mothers with gestational diabetes. The purpose of the study was to compare the metabolomic profile of infants of IGDMs to that of infants of healthy mothers to evaluate if differences remain despite a strict control of gestational diabetes. The metabolomic analysis was performed on serum from 30 term IGDMs and 40 controls. The results of this investigation showed that IGDMs have lower level of glucose and higher level of pyruvate, histidine, alanine, valine, methionine, arginine, lysine, hypoxanthine, lipoproteins, and lipids than controls, but showed no clinical differences. Authors concluded that prolonged fetal exposure to hyperglycemia during pregnancy can change neonatal metabolomic profile at birth without affecting the clinical course.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
