Metabolite Measurement: Pitfalls to Avoid and Practices to Follow.

Metabolites are the small biological molecules involved in energy conversion and biosynthesis. Studying metabolism is inherently challenging due to metabolites' reactivity, structural diversity, and broad concentration range. Herein, we review the common pitfalls encountered in metabolomics and provide concrete guidelines for obtaining accurate metabolite measurements, focusing on water-soluble primary metabolites. We show how seemingly straightforward sample preparation methods can introduce systematic errors (e.g., owing to interconversion among metabolites) and how proper selection of quenching solvent (e.g., acidic acetonitrile:methanol:water) can mitigate such problems. We discuss the specific strengths, pitfalls, and best practices for each common analytical platform: liquid chromatography-mass spectrometry (LC-MS), gas chromatography-mass spectrometry (GC-MS), nuclear magnetic resonance (NMR), and enzyme assays. Together this information provides a pragmatic knowledge base for carrying out biologically informative metabolite measurements.

Feature impact assessment: a new score to identify relevant metabolomics features in artificial neural networks using validated labels.

INTRODUCTION: Artificial Neural Networks (ANN) are increasingly used in metabolomics but are hard to interpret. OBJECTIVES: We aimed at developing a feature impact score that is model-agnostic, simple, and interpretable. METHODS: Feature Impact Assessment (FIA) is calculated by varying combinations of features within their observed value range and checking for changes in prediction outcomes. FIA was implemented in R and tested on metabolomics datasets. RESULTS: FIA exceeded LIME and SHAP in selecting biologically meaningful features. Values were comparable across different ANN architectures. CONCLUSION: FIA is a novel score ranking feature impact, helping interpreting ANN in the metabolomics field.

Affine Transformation of Negative Values for NMR Metabolomics Using the mrbin R Package.

Data from untargeted metabolomics studies employing nuclear magnetic resonance (NMR) spectroscopy oftentimes contain negative values. These negative values hamper data processing and analysis algorithms and prevent the use of such data in multiomics integration settings. New methods to deal with such negative values are thus an urgent need in the metabolomics community. This study presents affine transformation of negative values (ATNV), a novel algorithm for replacement of negative values in NMR data sets. ATNV was implemented in the R package mrbin, which features interactive menus for user-friendly application and is available for free for various operating systems within the free R statistical programming language. The novel algorithms were tested on a set of human urinary NMR spectra and were able to successfully identify relevant metabolites.

Maternal and Cord Blood Metabolite Associations with Gestational Weight Gain and Pregnancy Health Outcomes.

Pre-pregnancy obesity and excessive gestational weight gain (GWG) are risk factors for future maternal and childhood obesity. Maternal obesity is potentially communicated to the fetus in part by the metabolome, altering the child's metabolic program in early development. Fasting maternal blood samples from 37 singleton pregnancies at 25-28 weeks of gestation were obtained from mothers with pre-pregnancy body mass indexes (BMIs) between 18 and 40 kg/m2. Various health measures including GWG, diet, and physical activity were also assessed. At term (37-42 weeks), a venous umbilical cord sample was obtained. Serum metabolomic profiles were measured using nuclear magnetic resonance spectroscopy as well as a gut and metabolic hormone panel. Maternal and cord serum metabolites were tested for associations with pre-pregnancy BMI, GWG, health outcomes, and gut and metabolic hormones. While cord blood metabolites showed no significant correlation to maternal obesity status or other measured health outcomes, maternal serum metabolites showed distinct profiles for lean, overweight, and obese women. Additionally, four serum metabolites, namely, glutamate, lysine, pyruvate, and valine, allowed prediction of excessive GWG when pre-pregnancy BMI was controlled. Metabolic biomarkers predictive of GWG are reported and, if validated, could aid in the guidance of prenatal weight management plans as the majority of pregnancy weight gain occurs in the third trimester.

nmrML: A Community Supported Open Data Standard for the Description, Storage, and Exchange of NMR Data.

NMR is a widely used analytical technique with a growing number of repositories available. As a result, demands for a vendor-agnostic, open data format for long-term archiving of NMR data have emerged with the aim to ease and encourage sharing, comparison, and reuse of NMR data. Here we present nmrML, an open XML-based exchange and storage format for NMR spectral data. The nmrML format is intended to be fully compatible with existing NMR data for chemical, biochemical, and metabolomics experiments. nmrML can capture raw NMR data, spectral data acquisition parameters, and where available spectral metadata, such as chemical structures associated with spectral assignments. The nmrML format is compatible with pure-compound NMR data for reference spectral libraries as well as NMR data from complex biomixtures, i.e., metabolomics experiments. To facilitate format conversions, we provide nmrML converters for Bruker, JEOL and Agilent/Varian vendor formats. In addition, easy-to-use Web-based spectral viewing, processing, and spectral assignment tools that read and write nmrML have been developed. Software libraries and Web services for data validation are available for tool developers and end-users. The nmrML format has already been adopted for capturing and disseminating NMR data for small molecules by several open source data processing tools and metabolomics reference spectral libraries, e.g., serving as storage format for the MetaboLights data repository. The nmrML open access data standard has been endorsed by the Metabolomics Standards Initiative (MSI), and we here encourage user participation and feedback to increase usability and make it a successful standard.

Ketogenic diet leads to O-GlcNAc modification in the BTBRT+tf/j mouse model of autism.

BACKGROUND: Protein O-linked-ß-N-acetyl glucosamine (O-GlcNAc) is a post-translational modification to Ser/Thr residues that integrates energy supply with demand. Abnormal O-GlcNAc patterning is evident in several neurological disease states including epilepsy, Alzheimer's disease and autism spectrum disorder (ASD). A potential treatment option for these disorders includes the high-fat, low-carbohydrate, ketogenic diet (KD). The goal of this study was to determine whether the KD induces changes in O-GlcNAc in the BTBRT+tf/j (BTBR) mouse model of ASD. METHODS: Juvenile male (5weeks), age-matched C57 or BTBR mice consumed a chow diet (13% kcal fat) or KD (75% kcal fat) for 10-14days. Following these diets, brain (prefrontal cortex) and liver were examined for gene expression levels of key O-GlcNAc mediators, global and protein specific O-GlcNAc as well as indicators of energy status. RESULTS: The KD reduced global O-GlcNAc in the livers of all animals (p<0.05). Reductions were likely mediated by lower protein levels of O-GlcNAc transferase (OGT) and increased O-GlcNAcase (OGA) (p<0.05). In contrast, no differences in global O-GlcNAc were noted in the brain (p>0.05), yet OGT and OGA expression (mRNA) were elevated in both C57 and BTBR animals (p<0.05). CONCLUSIONS: The KD has tissue specific impacts on O-GlcNAc. Although levels of O-GlcNAc play an important role in neurodevelopment, levels of this modification in the juvenile mouse brain were stable with the KD despite large fluctuations in energy status. This suggests that it is unlikely that the KD exerts it therapeutic benefit in the BTBR model of ASD by O-GlcNAc related pathways.

Metabolomic Modeling To Monitor Host Responsiveness to Gut Microbiota Manipulation in the BTBR(T+tf/j) Mouse.

The microbiota, the entirety of microorganisms residing in the gut, is increasingly recognized as an environmental factor in the maintenance of health and the development of disease. The objective of this analysis was to model in vivo interactions between gut microbiota and both serum and liver metabolites. Different genotypic models (C57BL/6 and BTBR(T+tf/j) mice) were studied in combination with significant dietary manipulations (chow vs ketogenic diets) to perturb the gut microbiota. Diet rather than genotype was the primary driver of microbial changes, with the ketogenic diet diminishing total bacterial levels. Fecal but not cecal microbiota profiles were associated with the serum and liver metabolomes. Modeling metabolome-microbiota interactions showed fecal Clostridium leptum to have the greatest impact on host metabolism, significantly correlating with 10 circulating metabolites, including 5 metabolites that did not correlate with any other microbes. C. leptum correlated negatively with serum ketones and positively with glucose and glutamine. Interestingly, microbial groups most strongly correlated with host metabolism were those modulating gut barrier function, the primary site of microbe-host interactions. These results show very robust relationships and provide a basis for future work wherein the compositional and functional associations of the microbiome can be modeled in the context of the metabolome.

Polymorphisms within the APOBR gene are highly associated with milk levels of prognostic ketosis biomarkers in dairy cows.

Essentially all high-yielding dairy cows experience a negative energy balance during early lactation leading to increased lipomobilization, which is a normal physiological response. However, a severe energy deficit may lead to high levels of ketone bodies and, subsequently, to subclinical or clinical ketosis. It has previously been reported that the ratio of glycerophosphocholine to phosphocholine in milk is a prognostic biomarker for the risk of ketosis in dairy cattle. It was hypothesized that this ratio reflects the ability to break down blood phosphatidylcholine as a fatty acid resource. In the current study, 248 animals from a previous study were genotyped with Illumina BovineSNP50 BeadChip, and genome-wide association studies were carried out for the milk levels of phosphocholine, glycerophosphocholine, and the ratio of both metabolites. It was demonstrated that the latter two traits are heritable with h2 = 0.43 and h2 = 0.34, respectively. A major quantitative trait locus was identified on cattle chromosome 25. The APOBR gene, coding for the apolipoprotein B receptor, is located within this region and was analyzed as a candidate gene. The analysis revealed highly significant associations of polymorphisms within the gene with glycerophosphocholine as well as the metabolite ratio. These findings support the hypothesis that differences in the ability to take up blood phosphatidylcholine from low-density lipoproteins play an important role in early lactation metabolic stability of dairy cows and indicate APOBR to contain a causative variant.

Metabolomics reveals the sex-specific effects of the SORT1 low-density lipoprotein cholesterol locus in healthy young adults.

Metabolite profiles of individuals possessing either the cardiovascular risk or protective variants of the low-density lipoprotein cholesterol (LDL-C) associated 1p13.3 locus of the SORT1 gene (rs646776) were analyzed. Serum metabolites and lipids were assessed using LC-MS-based metabolomics in a healthy young population (n = 138: 95 males, 43 females). Although no significant differences were observed in the combined cohort, divergent sex effects were identified. Females carrying the protective allele showed increased phosphatidylcholines, very long chain fatty acids (>C20), and unsaturated fatty acids. Unsaturated fatty acids are considered to be protective against cardiovascular disease. In contrast, males carrying the protective allele exhibited decreased long-chain fatty acids (≤C20) and sphingomyelins, which is similarly considered to decrease cardiovascular disease risk. No significant changes in clinically assessed lipids such as LDL-C, high-density lipoprotein (HDL-C), total cholesterol, or triglycerides were observed in females, whereas only LDL-C was significantly changed in males. This indicates that, apart from reducing LDL-C, other mechanisms may contribute to the protective effect of the SORT1 locus. Thus, the analysis of metabolic biomarkers might reveal early disease development that may be overlooked by relying on standard clinical parameters.

Consuming royal jelly alters several phenotypes associated with overwintering dormancy in mosquitoes.

Introduction: Females of the Northern house mosquito, Culex pipiens, enter an overwintering dormancy, or diapause, in response to short day lengths and low environmental temperatures that is characterized by small egg follicles and high starvation resistance. During diapause, Culex pipiens Major Royal Jelly Protein 1 ortholog (CpMRJP1) is upregulated in females of Cx. pipiens. This protein is highly abundant in royal jelly, a substance produced by honey bees (Apis mellifera), that is fed to future queens throughout larval development and induces the queen phenotype (e.g., high reproductive activity and longer lifespan). However, the role of CpMRJP1 in Cx. pipiens is unknown. Methods: We first conducted a phylogenetic analysis to determine how the sequence of CpMRJP1 compares with other species. We then investigated how supplementing the diets of both diapausing and nondiapausing females of Cx. pipiens with royal jelly affects egg follicle length, fat content, protein content, starvation resistance, and metabolic profile. Results: We found that feeding royal jelly to females reared in long-day, diapause-averting conditions significantly reduced the egg follicle lengths and switched their metabolic profiles to be similar to diapausing females. In contrast, feeding royal jelly to females reared in short-day, diapause-inducing conditions significantly reduced lifespan and switched their metabolic profile to be similar nondiapausing mosquitoes. Moreover, RNAi directed against CpMRJPI significantly increased egg follicle length of short-day reared females, suggesting that these females averted diapause. Discussion: Taken together, our data show that consuming royal jelly reverses several key seasonal phenotypes of Cx. pipiens and that these responses are likely mediated in part by CpMRJP1.

Correlations between milk and plasma levels of amino and carboxylic acids in dairy cows.

The objective of this study was to investigate the relationship between the concentrations of 19 amino acids, glucose, and seven carboxylic acids in the blood and milk of dairy cows and their correlations with established markers of ketosis. To that end, blood plasma and milk specimens were collected throughout lactation in two breeds of dairy cows of different milk yield. Plasma concentrations of glucose, pyruvate, lactate, α-aminobutyrate, ß-hydroxybutyrate (BHBA), and most amino acids, except for glutamate and aspartate, were on average 9.9-fold higher than their respective milk levels. In contrast, glutamate, aspartate, and the Krebs cycle intermediates succinate, fumarate, malate, and citrate were on average 9.1-fold higher in milk than in plasma. For most metabolites, with the exception of BHBA and threonine, no significant correlations were observed between their levels in plasma and milk. Additionally, milk levels of acetone showed significant direct relationships with the glycine-to-alanine ratio and the BHBA concentration in plasma. The marked decline in plasma concentrations of glucose, pyruvate, lactate, and alanine in cows with plasma BHBA levels above the diagnostic cutoff point for subclinical ketosis suggests that these animals fail to meet their glucose demand and, as a consequence, rely increasingly on ketone bodies as a source of energy. The concomitant increase in plasma glycine may reflect not only the excessive depletion of protein reserves but also a potential deficiency of vitamin B6.

Early changes in the liver-soluble proteome from mice fed a nonalcoholic steatohepatitis inducing diet.

Despite the increasing incidence of nonalcoholic steatohepatitis (NASH) with the rise in lifestyle-related diseases such as the metabolic syndrome, little is known about the changes in the liver proteome that precede the onset of inflammation and fibrosis. Here, we investigated early changes in the liver-soluble proteome of female C57BL/6N mice fed an NASH-inducing diet by 2D-DIGE and nano-HPLC-MS/MS. In parallel, histology and measurements of hepatic content of triglycerides, cholesterol and intermediates of the methionine cycle were performed. Hepatic steatosis manifested itself after 2 days of feeding, albeit significant changes in the liver-soluble proteome were not evident before day 10 in the absence of inflammatory or fibrotic signs. Proteomic alterations affected mainly energy and amino acid metabolism, detoxification processes, urea cycle, and the one-carbon/S-adenosylmethionine pathways. Additionally, intermediates of relevant affected pathways were quantified from liver tissue, confirming the findings from the proteomic analysis.

Performance evaluation of algorithms for the classification of metabolic 1H NMR fingerprints.

Nontargeted metabolite fingerprinting is increasingly applied to biomedical classification. The choice of classification algorithm may have a considerable impact on outcome. In this study, employing nested cross-validation for assessing predictive performance, six binary classification algorithms in combination with different strategies for data-driven feature selection were systematically compared on five data sets of urine, serum, plasma, and milk one-dimensional fingerprints obtained by proton nuclear magnetic resonance (NMR) spectroscopy. Support Vector Machines and Random Forests combined with t-score-based feature filtering performed well on most data sets, whereas the performance of the other tested methods varied between data sets.

NMR metabolomic analysis of dairy cows reveals milk glycerophosphocholine to phosphocholine ratio as prognostic biomarker for risk of ketosis.

Ketosis is a common metabolic disease in dairy cows. Diagnostic markers for ketosis such as acetone and beta-hydroxybutyric acid (BHBA) are known, but disease prediction remains an unsolved challenge. Milk is a steadily available biofluid and routinely collected on a daily basis. This high availability makes milk superior to blood or urine samples for diagnostic purposes. In this contribution, we show that high milk glycerophosphocholine (GPC) levels and high ratios of GPC to phosphocholine (PC) allow for the reliable selection of healthy and metabolically stable cows for breeding purposes. Throughout lactation, high GPC values are connected with a low ketosis incidence. During the first month of lactation, molar GPC/PC ratios equal or greater than 2.5 indicate a very low risk for developing ketosis. This threshold was validated for different breeds (Holstein-Friesian, Brown Swiss, and Simmental Fleckvieh) and for animals in different lactations, with observed odds ratios between 1.5 and 2.38. In contrast to acetone and BHBA, these measures are independent of the acute disease status. A possible explanation for the predictive effect is that GPC and PC are measures for the ability to break down phospholipids as a fatty acid source to meet the enhanced energy requirements of early lactation.

Detection of autosomal dominant polycystic kidney disease by NMR spectroscopic fingerprinting of urine.

Autosomal dominant polycystic kidney disease (ADPKD) is a frequent cause of kidney failure; however, urinary biomarkers for the disease are lacking. In a step towards identifying such markers, we used multidimensional-multinuclear nuclear magnetic resonance (NMR) spectroscopy with support vector machine-based classification and analyzed urine specimens of 54 patients with ADPKD and slightly reduced estimated glomerular filtration rates. Within this cohort, 35 received medication for arterial hypertension and 19 did not. The results were compared with NMR profiles of 46 healthy volunteers, 10 ADPKD patients on hemodialysis with residual renal function, 16 kidney transplant patients, and 52 type 2 diabetic patients with chronic kidney disease. Based on the average of 51 out of 701 NMR features, we could reliably discriminate ADPKD patients with moderately advanced disease from ADPKD patients with end-stage renal disease, patients with chronic kidney disease of other etiologies, and healthy probands with an accuracy of >80%. Of the 35 patients with ADPKD receiving medication for hypertension, most showed increased excretion of proteins and also methanol. In contrast, elevated urinary methanol was not found in any of the control and other patient groups. Thus, we found that NMR fingerprinting of urine differentiates ADPKD from several other kidney diseases and individuals with normal kidney function. The diagnostic and prognostic potential of these profiles requires further evaluation.

Deep Learning for Rapid Identification of Microbes Using Metabolomics Profiles.

Rapid detection of viable microbes remains a challenge in fields such as microbial food safety. We here present the application of deep learning algorithms to the rapid detection of pathogenic and non-pathogenic microbes using metabolomics data. Microbes were incubated for 4 h in a protein-free defined medium, followed by 1D 1H nuclear magnetic resonance (NMR) spectroscopy measurements. NMR spectra were analyzed by spectral binning in an untargeted metabolomics approach. We trained multilayer ("deep") artificial neural networks (ANN) on the data and used the resulting models to predict spectra of unknown microbes. ANN predicted unknown microbes in this laboratory setting with an average accuracy of 99.2% when using a simple feature selection method. We also describe learning behavior of the employed ANN and the optimization strategies that worked well with these networks for our datasets. Performance was compared to other current data analysis methods, and ANN consistently scored higher than random forest models and support vector machines, highlighting the potential of deep learning in metabolomics data analysis.

Determining the effects of nutrition on the reproductive physiology of male mosquitoes.

Nutrition affects multiple aspects of insect physiology such as body size and fecundity, but we lack a detailed understanding of how nutrition influences the reproductive physiology of male insects such as mosquitoes. Given that female mosquitoes are vectors of many deadly diseases and can quickly proliferate, understanding how male nutrition impacts female fecundity could be of critical importance. To uncover the relationship between nutrition in adult male mosquitoes and its impacts on reproductive physiology, we reared larvae of the Northern house mosquito, Culex pipiens, on a standard lab diet and divided adult males among three different dietary treatments: low (3%), moderate (10%), and high (20%) sucrose. We found that although overall body size did not differ among treatments, one-week-old males raised on the 3% sucrose diet had significantly smaller male accessory glands (MAGs) compared to males that consumed the 10% and the 20% sucrose diets. Diet affected whole-body lipid content but did not affect whole-body protein content. Using nuclear magnetic resonance (NMR) spectroscopy, we found that diet altered the metabolic composition of the MAGs, including changes in lactic acid, formic acid, and glucose. We also observed changes in protein and lipid abundance and composition in MAGs. Females who mated with males on the 3% diet were found to produce significantly fewer larvae than females who had mated with males on the 10% diet. Taken together, our results demonstrate that the diet of adult male mosquitoes clearly affects male reproductive physiology and female fecundity.

Metabolomic Markers of Storage Temperature and Time in Pasteurized Milk.

The current date labeling system for pasteurized milk is based on the predicted growth of spoilage microorganisms, but inherent inaccuracies and the inability to account for environmental factors (e.g., temperature fluctuations) contribute to household and retail food waste. Improved shelf-life estimation can be achieved by monitoring milk quality in real-time. In this study, we identify and quantify metabolites changing over storage temperature and time, the main factors affecting milk stability. Pasteurized 2% fat milk was stored at 4, 10, 15, and 20 °C. Metabolite change was analyzed using untargeted and targeted nuclear magnetic resonance (NMR) metabolomics approaches. Several metabolites correlated significantly to storage time and temperature. Citric acid decreased linearly over time at a temperature-dependent rate. Ethanol, formic acid, acetic acid, lactic acid, and succinic acid increased non-linearly after an initial period of minimal increase. Butyric acid exhibited strong inverse temperature dependencies. This study provides the first analysis of the effect of time and temperature on the concentration of key metabolites during milk storage. Candidate molecules for shelf-life monitoring have been identified, and the results improve our understanding of molecular changes during milk storage. These results will inform the development of real-time shelf-life indicators for milk, helping to reduce milk waste.

Caffeine-Containing Energy Shots Cause Acute Impaired Glucoregulation in Adolescents.

Caffeine-containing, nutritionally fortified energy shots are consumed at high rates by adolescents, yet little is known about their metabolic impact. The purpose of this study was to examine the consequences of small format, caffeinated energy shots on glucose metabolism and gastrointestinal hormone secretion in adolescents. Twenty participants aged 13-19 years participated in a double-blind, randomized cross-over study consisting of two trials separated by 1-4 weeks. Participants consumed a volume-matched caffeinated energy shot (CAF, 5 mg/kg) or a decaffeinated energy shot (DECAF) followed by a 2 h oral glucose tolerance test. Blood samples were collected and area under the curve (AUC) calculated for glucose, insulin and gut and metabolic hormones. Consumption of CAF resulted in a 25% increase in glucose and a 26% increase in insulin area under the curve (AUC, p = 0.037; p < 0.0001) compared to DECAF. No impact on gut hormones was observed. To further characterize responses, individuals were classified as either slow or fast caffeine metabolizers based on an allele score. Glucose intolerance was greater in genetically fast vs. slow caffeine metabolizers and differences between groups were supported by distinct serum metabolomics separation. Consumption of caffeine-containing energy shots results in acute impaired glucoregulation in healthy adolescents as characterized by hyperinsulinemia following an oral glucose challenge.

Mesenchymal Stem Cells Shift Mitochondrial Dynamics and Enhance Oxidative Phosphorylation in Recipient Cells.

Mesenchymal stem cells (MSCs) are the most commonly used cells in tissue engineering and regenerative medicine. MSCs can promote host tissue repair through several different mechanisms including donor cell engraftment, release of cell signaling factors, and the transfer of healthy organelles to the host. In the present study, we examine the specific impacts of MSCs on mitochondrial morphology and function in host tissues. Employing in vitro cell culture of inherited mitochondrial disease and an in vivo animal experimental model of low-grade inflammation (high fat feeding), we show human-derived MSCs to alter mitochondrial function. MSC co-culture with skin fibroblasts from mitochondrial disease patients rescued aberrant mitochondrial morphology from a fission state to a more fused appearance indicating an effect of MSC co-culture on host cell mitochondrial network formation. In vivo experiments confirmed mitochondrial abundance and mitochondrial oxygen consumption rates were elevated in host tissues following MSC treatment. Furthermore, microarray profiling identified 226 genes with differential expression in the liver of animals treated with MSC, with cellular signaling, and actin cytoskeleton regulation as key upregulated processes. Collectively, our data indicate that MSC therapy rescues impaired mitochondrial morphology, enhances host metabolic capacity, and induces widespread host gene shifting. These results highlight the potential of MSCs to modulate mitochondria in both inherited and pathological disease states.