Metabolomics in the Developing Human Being

Metabolomics is based on the detailed analysis of metabolites and represents a unique chemical fingerprint of an organism. This approach allows assessing the dynamic behavior of biologic systems with multiple network interactions among individual components. The field of metabolic profiling has rapidly developed over the last decade, with successful applications in various research areas including toxicology, disease diagnosis and classification, pharmacology, and nutrition. This article provides a comprehensive account of existing data in the literature from animal and clinical studies on the use of metabolomics for improved understanding of medical conditions affecting the neonate and the developing human being.

  • The first results on metabolomics are available in perinatology and pediatrics and can help in defining an atlas of metabolic alterations in different conditions and pathologies.

  • Metabolomics could rapidly become mainstream in diagnosing and subsequently managing pathologic conditions.

  • The availability of information through the analysis of noninvasively collected fluid, such as urine, makes it extremely appealing.

  • For neonatology in particular the extended and specific amount of data generated by metabolomics could allow personalized nutritional and therapeutic interventions.

  • This approach blazes a revolutionary trail from reductionist medicine to holistic medicine, from descriptive medicine to predictive medicine, and from an epidemiologic perspective to a personalized approach.

Key Points

Introduction

The “omics” technologies (genomics, transcriptomics, proteomics, and metabolomics) have impacted the life sciences considerably over the last decade. Starting with genomics at the DNA level, transcriptomics at the RNA level, proteomics at the protein level, and metabolomics at the metabolite level, these scientific technologies are based on the postgenomic activity, assess biologic function at the level of cellular organization, and offer unique insights into small molecule regulation and signaling. It is well known that all cells of the human body are in constant and variable communication with the fluid compartments of the body. This communication helps the cell metabolites, peptides, and proteins to be part of a dynamic process in which they are either released from cells or taken up by cells from bodily fluids by a variety of mechanisms: normal excretion, transmembrane diffusion, or transport ( Fig. 1 ). At least to a certain extent, the biochemical and protein-based changes occurring within cells and organs are reflected in bodily fluids.

Fig. 1
Dynamic interactions of the several pathways.
( Adapted from Christians U, Albuisson J, Klawitter J. The role of metabolomics in the study of kidney diseases and in the development of diagnostic tools. In: Edelstein CL, editor. Biomarkers of kidney disease. 1st edition. London: Elsevier Inc; 2011; with permission.)

Metabolomics is not a novel approach to health and disease; one can find the basic idea of “omics” already appreciated by ancient Greeks who believed that changes in tissue and biologic fluids were early signs of pathology, and thus were capable of serving as indicators of disease processes. In 1506 Ullrich Pinder, in his book Epiphanie Medicorum , describes the possible medical value of different colors, smells, and taste of urine. The “omics” technologies allow clinicians to study what causes these different smells and colors as they discover that the differences represent quite complex changes in the biologic system. Aristotle wrote that “the whole is not represented by the sum of its components.” In our setting this means that the behavior of complex systems (systems biology) cannot be predicted from the properties of its individual components.

Definition

The study of metabolites in biologic systems, referred to as “metabolomics,” primarily involves the study of metabolism. The word “metabolism” derives from the Ancient Greek metabole , meaning “change.” This new “omic” discipline is based on the detailed analysis of metabolites, which are all the endogenous intracellular and extracellular compounds produced by the organism. In other terms, the advent of metabolomics, in which all of the metabolites in a given tissue or biofluid are examined, reveals the personal metabolic identity and the metabolites found in biologic fluids are “a snapshot of the chemical fingerprints that specific cellular processes leave behind.”

Metabolomics can also be defined as the study of the complete set of metabolites of low or intermediate molecular weight (including amino acids, organic acids, sugars, fatty acids, lipids, steroids, small peptides, vitamins, and so forth), reflecting physiology, developmental, or pathologic state of cell, tissue, organ, or organism. “Metabolome” refers to the complete set of all metabolites of an organism or system, whereas the “genome” refers to the complete set of a system’s genes. As such, metabolomics provides the opportunity to directly or indirectly assess molecular mechanisms through molecular markers or biomarkers. A biomarker is defined as “a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes or pharmacologic responses to therapeutic intervention.” It may be part of a single molecular entity but it can also be a group of several molecular entities, as in a molecular pattern or fingerprint. To shed light on this novel science called metabolomics, investigators claim that the metabolome is a quantitative descriptor of all endogenous low-molecular-weight components in a biologic sample, such as urine or plasma. Each cell type and biologic fluid has a characteristic set of metabolites that reflects the organism under particular environmental conditions and that fluctuates according to physiologic demands. The same investigators explain the difference between metabolomics and metabonomics and report that metabolomics is “the comprehensive quantitative analysis of all the metabolites of an organism or a specific biologic sample,” whereas metabonomics are “the quantitative measurement over time of the metabolic responses of an individual or population to a disease, drug, treatment or other challenge.” Metabolomics has to be distinguished from the common pharmacologic use of the term “metabolite” for the degradation of a drug or conjugation products of pharmaceutical agents or xenobiotica.

From 2009 to 2011, more than 5000 manuscripts have been published on this topic, with a dramatic increase in comparison with the previous years (2000–2003, 125 papers; 2004–2006, about 1000 papers; 2006–2008, about 2700 papers).

Definition

The study of metabolites in biologic systems, referred to as “metabolomics,” primarily involves the study of metabolism. The word “metabolism” derives from the Ancient Greek metabole , meaning “change.” This new “omic” discipline is based on the detailed analysis of metabolites, which are all the endogenous intracellular and extracellular compounds produced by the organism. In other terms, the advent of metabolomics, in which all of the metabolites in a given tissue or biofluid are examined, reveals the personal metabolic identity and the metabolites found in biologic fluids are “a snapshot of the chemical fingerprints that specific cellular processes leave behind.”

Metabolomics can also be defined as the study of the complete set of metabolites of low or intermediate molecular weight (including amino acids, organic acids, sugars, fatty acids, lipids, steroids, small peptides, vitamins, and so forth), reflecting physiology, developmental, or pathologic state of cell, tissue, organ, or organism. “Metabolome” refers to the complete set of all metabolites of an organism or system, whereas the “genome” refers to the complete set of a system’s genes. As such, metabolomics provides the opportunity to directly or indirectly assess molecular mechanisms through molecular markers or biomarkers. A biomarker is defined as “a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes or pharmacologic responses to therapeutic intervention.” It may be part of a single molecular entity but it can also be a group of several molecular entities, as in a molecular pattern or fingerprint. To shed light on this novel science called metabolomics, investigators claim that the metabolome is a quantitative descriptor of all endogenous low-molecular-weight components in a biologic sample, such as urine or plasma. Each cell type and biologic fluid has a characteristic set of metabolites that reflects the organism under particular environmental conditions and that fluctuates according to physiologic demands. The same investigators explain the difference between metabolomics and metabonomics and report that metabolomics is “the comprehensive quantitative analysis of all the metabolites of an organism or a specific biologic sample,” whereas metabonomics are “the quantitative measurement over time of the metabolic responses of an individual or population to a disease, drug, treatment or other challenge.” Metabolomics has to be distinguished from the common pharmacologic use of the term “metabolite” for the degradation of a drug or conjugation products of pharmaceutical agents or xenobiotica.

From 2009 to 2011, more than 5000 manuscripts have been published on this topic, with a dramatic increase in comparison with the previous years (2000–2003, 125 papers; 2004–2006, about 1000 papers; 2006–2008, about 2700 papers).

Advantages of this “omics” technology

The genotype of a patient can be substantially considered a static parameter and defines the risk or probability of reacting to a disease, drug, or environmental challenge in a certain way. However, the phenotype more closely reflects clinical reality at a given moment. The “omics” studies aim at shedding light on the phenotype of a cell, tissue, or organism through the full spectrum of metabolites present and allow to combine constitutive aspects (eg, genes) and environmental factors (eg, diet, xenobiotics, and drugs). A critical challenge for biology and medicine is to achieve a more open-minded approach of the complex cellular systems as a whole. Thus metabolomics reflect a dynamic behavior of biologic systems with multiple networks of interactions among individual components. Moreover, endogenous metabolites of a biofluid can describe the cellular phenotype, which in response to endogenous or exogenous stimuli, such as hormones, drugs, food, and chemicals, may induce biologic reactions, connecting genomics, transcriptomics, and proteomics and thus changing completely the way of studying and interpreting some cellular processes of diseases. Not long ago, the scientific world was mainly interested in understanding disease complexity by analyzing the significance of a single part or observing the microscopic world with the firm belief that genes are the only powerful discovery tools (eg, DNA analysis at gene level). By contrast, metabolomics describes the metabolic status of the human being and represents an open system influenced by either genotype–phenotype or genotype–environment interactions.

Metabolomic studies are typically performed in biofluids (urine, plasma, cerebrospinal fluid [CSF], maternal milk, or saliva) or tissues (kidney, liver, gut, or brain). Urine contains important information on the overall metabolic state of an individual and has a unique biochemical composition that is affected by multiple parameters, such as genotype, physiologic and pathologic conditions, environment, nutrition, drugs, and so forth. Urine is easy to collect, using simple, noninvasive or painless techniques. In neonates it can be collected using a cotton ball or by the “catch up” technique. Detailed analysis of the urinary metabolite patterns may help predict clinical presentations, allow for early detection of certain conditions and help monitor a patient’s disease progression, treatment response and/or organ recovery.

Understanding the biochemical and pathophysiologic mechanisms underlying the response of a biologic system to endogenous or exogenous stimuli may be the first step required to predict the type of response (in terms of efficacy and safety) of an organism to a specific pharmacologic or nonpharmacologic intervention, thus leading closer to providing “personalized medicine.” The development of individualized treatment regimens, optimized for the metabolic status of the patient, is a major goal of medicine in general and of neonatology in particular in the twenty-first century. The right therapy for the right patient, minimizing the side effects and maximizing the benefits, is the dream of every physician.

Metabolomics has the advantage of a limited number of variables (molecules being measured) compared with the approximately 25,000 genes tested in genomics, the approximately 85,000 transcripts assessed in transcriptomics, and the more than 10,000,000 proteins measured in proteomics. Metabolomics monitors only 1400 to 3000 metabolites and yet provides detailed information about the metabolic activity, reflecting the downstream changes in the genome, transcriptome, and proteome of an individual. The relatively small number of variables may be one of the major advantages of metabolomics compared with other “omics” sciences. Most of these metabolites participate in specific biochemical pathways, such as glycolysis, Krebs cycle, and lipid or amino acid metabolism; signal pathways, such as transmitters and hormones; and specific pathobiochemical processes, such as oxidative stress. Moreover, changes in specific metabolite patterns may reflect changes in real-time pathways and processes, thus making it easy to detect derangements that might have occurred and pointing out certain metabolic reactions.

The diverse composition of metabolites provides a wide range of physicochemical characteristics including molecular weight, hydrophobicity/hydrophilicity, acidity/basicity, and boiling point. The range of molecular weight, from 1 amu (proton) to greater than 1500 amu (eg, gangliosides, lipids, and small peptides), is significantly lower than that observed for proteins, transcripts, and genes. Hydrophobicity/hydrophilicity ranges from polar metabolites, such as low-molecular-weight amino acids to high-molecular-weight nonpolar lipids. Volatility ranges from low boiling point metabolites present in breath including isoprene and carbon dioxide, to high-molecular-weight lipids. The complex interaction between these diverse components is extremely challenging and covers a wide range of biochemical reactions of the human body.

Finally, metabolomics studies are readily available, because most large academic centers have expertise in analytical chemistry, statistics, and bioinformatics. As in all “omics” sciences, metabolomics is a multidisciplinary science involving clinicians, cell biologists, analytical chemists, clinical pharmacologists, biochemists, and statisticians.

How a study on metabolomics is designed

Collection of Samples

Mammalian biologic samples are complex and contain metabolites and low- and high-concentration matrix components. Immediately after collection of tissues, cells, urine, or other fluids, the temperature is reduced to subzero levels and samples stored at −80°C. Urine from healthy mammals has very low protein content and preparation steps are simple and normally involve dilution and analysis. Blood requires an extra step of preparation to allow separation of serum or plasma at 4°C before freezing and storage. The most complex and experimentally difficult systems to extract for metabolomics analysis are tissues; usually more than 30 mg of tissue is required. Homogenization and mechanical or chemical lysis of cell walls is required to allow the release of metabolites.

Metabolic Pattern Analysis

Two important steps are required in planning an “omics” protocol: experimental technique and multivariate data analysis. The study design (adequate sample size and the statistical analysis strategy) should be carefully planned to detect small differences in metabolic profiles in a population with wide biologic variation. As soon as the biofluids have been collected and prepared with specific attention to uniform methods, nuclear magnetic resonance (NMR) spectroscopy, gas chromatography–mass spectrometry, and liquid chromatography–mass spectrometry are the main experimental tools used to measure global sets of metabolites in biologic samples. Targeted (a priori) and untargeted (a posteriori) methods can be used with metabolomics. The first is hypothesis testing; the second, more innovative, is hypothesis generating.

Two distinct methods for metabolomic analysis exist at present: the first one is global metabolomics, which refers to true biomarker discovery and evaluates known or unknown biomarkers indicative of intracellular metabolism; the second method, metabolic fingerprinting, reflects the ways intracellular metabolism affects the external environment by consumption and secretion of metabolites. In both methods, all results are validated using samples collected from different patient cohorts, using independent validation data sets.

After data collection, raw data are converted into biologic knowledge. This requires rigorous biostatistics. Univariate analyses, such as t test, analysis of variance, Mann-Whitney U test, Wilcoxon signed-rank test, and logistic regression, are applied to identify metabolites as potential biomarkers that are capable of differentiating between groups (eg, patients with vs those without a certain disease). These univariate tests are typically used individually to globally screen all of the measured metabolites for an association with a disease, which is where the multiple comparison problems may potentially arise. In the absence of correction for the number of comparisons, testing of multiple hypotheses causes an increase in the probability of obtaining statistically false-positive results. Several statistical methods exist for correction of multiple testing, including the Bonferroni correction, Benjamin-Hochberg false discovery rate, and the Westfall-Young method. When strong evidence on a subset of metabolites is found, further testing is performed to validate their plausibility in target-specific studies, often with more detailed phenotypes (eg, subtypes or stages of a disease). Multivariate analyses are important in metabolomics studies because they can reduce the variability of the data and combine complex interactions, because multiple biomarkers rather than one by itself are more likely to be specific for interpretation of a derangement.

Among hundreds of metabolites, variables influencing the projection are selected. Metabolomics is focused to find the key metabolites that are able to characterize pathology; response to a therapeutic intervention; or nutritional modification (scale-free networks). In a simple way it can be considered like a bar code. Modern analytical technologies allow for the identification of patterns that confer significantly more information than the measurement of a single parameter, much as a bar code contains more information than a single number. When a metabolomics protocol is designed a clear research goal should be set, and a quantitative hypothesis to be tested statistically should be identified. Due to the enormous amount of generated data, complex bioinformatic tools, experts in statistics, biochemistry, physics and biotechnology are required.

Clinical applications

Metabolomics: New Biomarkers of Prenatal Health

Metabolomic studies in obstetrics and gynecology are not a novel concept. Since the 1960s, studies have reported the important role of specific metabolites in the dynamic interactions among fetus, placenta, and mother. Several studies highlight the strong correlation of biomarkers found in amniotic fluid, in the placenta, or other biofluids of pregnant mothers with fetal malformations (FM), preterm delivery (PTD), premature rupture of membranes (PROM), prediagnostic gestational diabetes mellitus (GDM), and preeclampsia (PE). In a recent paper by Diaz and coworkers, plasma and urine samples were collected at amniocentesis and metabolites’ association with prenatal disorders was analyzed. All pregnant women were more than 35 years old or had other medical conditions necessitating amniocentesis. One hundred and ninety-eight plasma and urine samples (only 61 subjects consented for donation of urine and plasma) were collected in total and NMR spectroscopy was applied for the experimental signaling of metabolites. In the FM group, increase in glucogenic amino acids valine and isoleucine and threonine in urine and plasma was reported. Other metabolites, cis -aconite, an intermediate of the tricarboxylic acid cycle (TCA), and hypoxanthine indicating ATP degradation were found in high levels in both biofluids; both metabolites are indicative of fetal stress. Finally, choline, which plays an important role in homocysteine metabolism, is highly excreted in FM maternal urine, whereas two other biomarkers N -methyl-2-pyridone-5-carboxamide and N -methylnicotinamide, products of abnormal nucleotide metabolism, are similarly found in large amounts in the urine samples of the FM group. The latter urine metabolites are also indicative of other prenatal disorders, such as PE, hypoxia, chromosomal disorders, and PTD (choline) or prediagnostic GDM ( N -methyl-2-pyridone-5-carboxamide, N -methylnicotinamide). Maternal plasma in the FM group showed lower plasma betaine and trimethylamine- N -oxide concentrations. Other investigators reported that amniotic fluid is a reliable biofluid for detection of multiple prenatal disorders. Amniotic fluid samples were obtained from pregnant women aged 13 to 42 years undergoing amniocentesis during the second trimester of pregnancy (14–25 weeks of gestation). Samples were classified in six groups: (1) the healthy pregnancy; (2) pre-PTD (women who gave birth before 37 weeks of gestation); (3) prediagnostic GDM; (4) FM; (5) PROM; and (6) chromosomal disorders. For the disorders studied, amniotic fluid biomarkers seemed to have the best predictive value for FM. Metabolites analysis showed that malformed fetuses have large needs in glucose, whereas glucolysis is enhanced resulting in glucose and lactate increases. The decrease in the glucogenic amino acids alanine, isoleucine, glutamate, phenylalanine, tyrosine, and valine may be an indication of their enhanced preferential use in gluconeogenesis compared with other amino acids, such as glycine, glutamine, serine, and threonine. A high glutamine level reflects kidney disorders, whereas increased glycine and serine suggest a disturbance in choline and amino acid metabolism. Decrease in leucine and R-oxoisovalerate and the marked increase in ascorbate are indicative of abnormal amino acid biosynthesis in the FM group. For the prediagnostic GDM group, amniotic fluid analysis reported an increase in glucose and a decrease in acetate, creatinine, formate, glutamate, glycine, proline, serine, taurine, formate, and creatinine, suggesting changes in amino acids biosynthesis. In the pre-PTD group, the marked increase in allantoin, a marker of oxidative stress, and a decrease in myo -inositol, promoter of fetal lung maturation, was reported.

Preeclampsia

Metabolic profiling has revealed that lipid and ketone body concentrations are lower in women with PE. Kenny and colleagues performed a two-phase study on the metabolomic signature of PE. In the first phase, plasma samples were collected at 15 ± 1 weeks of gestation from women who subsequently developed PE during their pregnancy. In the second phase, the results were validated with the findings of plasma samples of a different group with similar characteristics. An overlap of 14 metabolites was found; thus, a significant metabolic fingerprint was reported to be relevant to the early prediction of subsequent PE. Odibo and colleagues also highlighted the importance of PE-specific metabolomic signature. A prospective cohort of pregnant women who subsequently developed PE and a control group with normal pregnancy outcome were followed-up from the first-trimester of pregnancy to delivery. Maternal blood was obtained at 11 to 14 weeks of gestation and 40 acylcarnitine species (C2–C18 saturated, unsaturated, and hydroxylated) and 32 amino acids were analyzed by liquid chromatography–tandem mass spectrometry. Four metabolites (hydroxyhexanoylcarnitine, alanine, phenylalanine, and glutamate) were significantly higher in women who subsequently developed PE, findings that highlight the first trimester “omics” role in the early diagnosis and prediction of PE.

A recent paper by Heazell and coworkers reviewed the “omics” response of the normal and preeclamptic placental tissue to different oxygen levels. They demonstrated that placental tissue from uncomplicated pregnancies cultured in 1% oxygen (hypoxia) had metabolic similarities to explants from PE pregnancies cultured at 6% oxygen (normoxia). Nevertheless, the investigators suggested that more light should be shed on certain metabolites, such as lipids, glutamate, and glutamine, and metabolites related to tryptophan, leukotriene, and prostaglandin, because they too may play an important role in the metabolic profile of PE.

Horgan and coworkers in a similar study compared placental features seen in small for gestational age (SGA) cases and controls. Placental tissue from both groups was cultured in 1% (hypoxic), 6% (normoxic), and 20% (hyperoxic) oxygen. Metabolic footprints were analyzed and 574 metabolites showed significant difference between SGA and normal pregnancies at one or more oxygen concentrations. SGA explant media cultured under hypoxic conditions was observed, on a univariate level, to exhibit the same metabolic signature as controls cultured under normoxic conditions in 49% of the metabolites of interest, suggesting that SGA tissue is adjusted to hypoxic conditions in vivo. Glycerophospholipid and tryptophan metabolism were highlighted as areas of particular interest.

Two recent cross-sectional metabolomics studies reported the importance of an amniotic fluid metabolic signature for preterm labor (PL) with or without intra-amniotic infection or inflammation (IAI). More specifically, two large groups of women with spontaneous PL and intact membranes were followed retrospectively. Amniotic fluids were collected and classified in three smaller groups based on the pregnancy outcome: (1) PL but delivery at term, (2) PL with PTD but no IAI, and (3) PL with IAI. Amniotic fluid biomarkers successfully predicted the pregnancy outcome in both groups with PL and with or without IAI. Interestingly, a decrease in amniotic fluid carbohydrates was associated with PTD with or without IAI, whereas an increase in amino acid metabolites is a unique feature of PTL with IAI. The opposite was true in patients with PL who delivered at term where carbohydrates, such as mannose, galactose, and fructose, were found in small amounts in the amniotic fluid, whereas amino acids, such as alanine, glutamine, and glutamic acid, were detected in low levels. This altered amniotic fluid composition was mainly explained by the presence of bacteria, which use carbohydrates as nutrients, and the catabolic state of the septic fetus. Two metabolites, methyladenine and diamino pimelic acid, components of bacterial processes and bacterial wall, respectively, were highlighted as the most significant biomarkers in this setting, in group classification.

Metabolomics in Nonhuman Models

Metabolomics analysis performed in nonhuman models also can shed light on the metabolic profile of the fetus and neonate with certain clinical conditions. A recent paper highlighted the plasma metabolome of fetal sheep brain after inflammatory-induced exposure. The fetal sheep were injected with Escherichia coli lipopolysaccharide (LPS) or saline injection by the umbilical vein and subsequently fetal blood was collected at specified time intervals. Postmortem, the white and gray matter were assessed. Within the first 3 days, LPS exposure caused hypoxia and a significant increase of the Krebs cycle intermediates, and alanine and lactate, and a subsequent decrease in hexoses; oxysteroles (24-hydroxycholesterol-24OHC, 25-hydroxycholesterol-25OHC), 12S-HETE, and spermidine increased after LPS administration, peaking at 6 hours. Moreover, 6 to 9 days later there was a delayed opposite effect of LPS on energy metabolites, hyperoxia, and elevation of sphingomyelins, kynurenine, 3-hydroxykynurenine, putrescine, and asymmetric dimethylarginine (ADMA), the latter ones known to be responsible markers and mediators of inflammation. The role of oxysteroles in proapoptotic pathways of the white matter and the role of ADMA in the regulation of microvascular tone and the endothelial function were highlighted as the metabolic profile of the relevant brain injury.

Metabolic studies in newborn piglets, which underwent hypoxia and then reoxygenation with different oxygen concentrations, revealed the strong association of different resuscitation patterns with metabolic changes. Three different groups of hypoxic newborn piglets received, respectively, 100% oxygen for 60 minutes, 21% oxygen for 60 minutes, 100% oxygen for 15 minutes, and then 21% oxygen for 45 minutes. Plasma parameters (eg, lactate, low pH, and base deficit) were not correlated with the duration of hypoxia; prolonged hypoxia revealed a low level of free and total carnitine, required for normal mitochondrial function, cellular antioxidant activity, and an increase in long-chain acylcarnitines, which have toxic potential. Reoxygenation with different oxygen concentrations showed that hyperoxia was associated with a slower decline of Krebs cycle intermediates and an increase in lanosterole (product of ineffective cholesterol synthesis) and oxysteroles, both indicative of acute neuronal damage.

Atzori and coworkers have recently described metabolomics in urine samples from newborn piglets undergoing hypoxia followed by resuscitation with different oxygen concentrations (ranging from 18% to 100%). Despite reoxygenation, 7 out of 10 piglets became asystolic and died. The most significant urine metabolites, which were different between these discriminating groups, were urea, creatinine, malonate, methylguanidine, and hydroxyisobutyric acid. Malonate, urea, and creatinine are known metabolites in neurologic and renal disorders, aerobic metabolism, and cell death. Metabolomics may predict mortality postasphyxia.

Beckstrom and coworkers reported the role of metabolites in perinatal asphyxia in another nonhuman model. After hysterectomy, umbilical cord was clamped in a Macaca nemestrina model and blood was collected before and 15 or 18 minutes after cord occlusion followed by postnatal sampling at 5 minutes of age. Metabolomics analysis with comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry method revealed 50 metabolites with the greatest change preasphyxia to postasphyxia. Fifteen of the 50 metabolites were significantly elevated in response to asphyxia, 10 of which remained significantly different compared with control animals. Similarly, lactate and creatinine were strongly identified as biomarkers used clinically to assess the degree of hypoxic-ischemic injury caused by perinatal asphyxia and new metabolites including succinic acid and malate (intermediates in the Krebs cycle) and arachidonic acid (a brain fatty acid and inflammatory marker) were increased, implying a disruption in metabolic pathways.

NMR-based metabolomics analysis was performed in an experimental model of septic rats versus control group. Sepsis was induced by cecal ligation and puncture and lung tissue, bronchoalveolar lavage (BAL) fluid, and serum samples were obtained. The increase in alanine concentration in lung tissue and serum samples of septic rats resulted from enhanced pyruvate metabolism and transamination to alanine by the Cori cycle as previously shown in sepsis. Creatine, which was found elevated in lung tissue, BAL, and serum, is a nitrogenous organic acid involved in inflammatory responses. An increase in phosphoethanolamine in the serum of septic rats could indicate cell damage and phospholipid degradation. The increase in serum acetoacetate might be related to enhanced fatty acid oxidation in septic rats. The decreased serum level of formate in septic rats indicates an increased biosynthesis of purine nucleotides in sepsis. Myo -inositol, which was found in high levels in lung tissue, but in low concentrations in the BAL, plays critical roles in endotoxin-induced vascular smooth muscle hypocontractility.

A similar nonhuman model study reported the metabolic changes during birth transition. Postterm primates gave birth by hysterotomy and prebirth blood samples followed by eight time-points postbirth samples were obtained and analyzed. One hundred metabolites were identified during this transition. Of these 100 metabolites, 23 demonstrated significant change during the first 72 hours postbirth. Of note, four intermediates of the TCA cycle (α-ketoglutaric acid, fumaric acid, malic acid, and succinyl-CoA) were found elevated, which is in accordance with the transition of the neonate from a hypoxic environment in utero to an oxygen-rich one postbirth. Myo -inositol and glutaminic acid, which are both correlated with hypoxic ischemic encephalopathy, were similarly elevated in serum samples. Myo -inositol is a precursor for inositol phospholipids and glutamic acid is involved in metabolic pathways, such as urea cycle, or serves as a central nervous system signaling molecule for neuronal apoptosis.

Liu and coworkers have recently highlighted the metabolomics application in an experimental protocol of brain injury and hypothermia. Neonatal rat brain slices were divided into three groups, and in the first group 45-minute oxygen-glucose deprivation with a 3 hourly mild hypothermia (32°C) was applied. In the second group, oxygen-glucose deprivation was followed by hypothermia after a 15-minute delay. Total normothermia (37°C) was applied in the control group. Hypothermia was followed by a 3-hour normothermic recovery. “Omics” analysis reported that the final ATP levels, severely decreased at normothermia, equally recovered by immediate and delayed hypothermia and cell death was greater with normothermia and delayed hypothermia, compared with immediate hypothermia; thus, the two hypothermia-treated groups totally restored their initial high-energy phosphates, whereas large differences in early cell death were observed. This implies that the ATP levels do not always reflect the cellular function, and more data are needed to direct optimal cooling temperatures, duration, and rewarming regimens.

In an experimental study in newborn rats, a gentamicin-induced nephrotoxicity was associated with a distinct pattern of urinary metabolites. In particular, 14 parameters were significantly modified by gentamicin administration, including glucose, galactose, N -acetylglucosamine, myo -inositol, butanoic acid, and 3-hydroxybutyrate, all which were increased about threefold, and citrulline, pseudouridine, which elevated at lower levels.

In another study preterm pigs, used as infant models, were given control treatment or broad-spectrum antibiotics just after birth by cesarean section. A close link between the antibiotic treatment, the presence of necrotizing enterocolitis, and the identified urinary metabolites was described, suggesting that urine metabolome could serve as an early biomarker for and subsequent progression of necrotizing enterocolitis in preterm neonates.

Neonatology

“Omics” studies in neonatology suggest that metabolic profiling can indeed play an important role in detecting multiple diseases of the neonatal period. Investigators report the variable application of metabolomics, using biofluids with noninvasive techniques, mainly urine sampling. Dessì and colleagues have recently reported results on metabolomics analysis in urine samples collected from intrauterine growth restricted and appropriate for gestational age babies on the first and fourth day of life. Alterations of three metabolic pathways were highlighted: (1) arginine and proline; (2) the urea cycle; and (3) glycine, serine, and threonine. Urine creatinine and myo -inositol were in high levels, the latter one representing a low fetal insulin production. Taken together all these observations indicate that the major effects of intrauterine growth restriction are on the brain and kidney and are associated with metabolites further involved in the metabolic syndrome. The same investigators described a distinct metabolic profile depending on the type of delivery. Twenty newborns delivered by caesarian section or spontaneous labor underwent metabolomics analysis; their metabolic fingerprint was quite different. Allantoin, betaine, and glycine were significantly higher in newborns born spontaneously than those born by caesarean section. A similar study in piglets showed that piglets born after a caesarian section had a metabolomics profile indicative of hepatic steatosis.

Data from our group suggest that metabolomics may predict the postmaturation of preterm and term neonates. The differences in urine metabolites at birth (mainly tyrosine metabolism, tyrosine, tryptophan, phenylalanine biosynthesis, urea cycle, arginine, and proline metabolism) reveal that gestational age has a strong effect on the metabolic profile of the neonate ( Fig. 2 ). Similarly, in a small study including newborns with respiratory distress syndrome, BAL fluid was obtained before and after surfactant administration, during mechanical ventilation, and at each extubation time point. Metabolomics analysis showed that 10 (undecane, decanoic acid, dodecanoic acid, hexadecanoic acid, octadecanoic acid, hexadecanoic acid methyl ester, 9-octadecanoic acid, tetracosanoic acid, myristic acid, and phosphate) out of 25 metabolites were overexpressed in neonates who required ventilation after surfactant treatment. Thus, metabolomics profiling of the BAL fluid may be a promising diagnostic tool for future management of neonates with respiratory distress syndrome.

Oct 3, 2017 | Posted by in PEDIATRICS | Comments Off on Metabolomics in the Developing Human Being

Full access? Get Clinical Tree

Get Clinical Tree app for offline access