Unveiling cellular biochemical reactions via metabolomics-driven approaches

Metabolomics-based development of bioproduction processes toward industrial-scale production

Microbial biomanufacturing offers a promising, environment-friendly platform for next-generation chemical production. However, its limited industrial implementation is attributed to the slow production rates of target compounds and the time-intensive engineering of high-yield strains. This review highlights how metabolomics expedites bioproduction development, as demonstrated through case studies of its integration into microbial strain engineering, culture optimization, and model construction. The Design-Build-Test-Learn (DBTL) cycle serves as a standard workflow for strain engineering. Process development, including the optimization of culture conditions and scale-up, is crucial for industrial production. In silico models facilitate the development of strains and processes. Metabolomics is a powerful driver of the DBTL framework, process development, and model construction.

Inducible nitric oxide synthase activity mediates TNF-α-induced endothelial cell dysfunction

Inducible nitric oxide synthase (iNOS) and vascular endothelial dysfunction have been implicated in the development and progression of atherosclerosis. This study aimed to elucidate the role of iNOS in vascular endothelial dysfunction. Ultrahigh performance liquid chromatography-quadrupole time-of-flight mass spectrometry combined with multivariate data analysis was used to characterize the metabolic changes in human umbilical vein endothelial cells (HUVECs) in response to different treatment conditions. In addition, molecular biology techniques were employed to explain the molecular mechanisms underlying the role of iNOS in vascular endothelial dysfunction. Tumor necrosis factor-α (TNF-α) enhances the expression of iNOS, TXNIP, and the level of reactive oxygen species (ROS) facilitates the entry of nuclear factor-κB (NF-κB) into the nucleus and promotes injury in HUVECs. iNOS deficiency reversed the TNF-α-mediated pathological changes in HUVECs. Moreover, TNF-α increased the expression of tumor necrosis factor receptor-2 (TNFR-2) and the levels of p-IκBα and IL-6 proteins and CD31, ICAM-1, and VCAM-1 protein expression, which was significantly reduced in HUVECs with iNOS deficiency. In addition, treating HUVECs in the absence or presence of TNF-α or iNOS, respectively, enabled the identification of putative endogenous biomarkers associated with endothelial dysfunction. These biomarkers were involved in critical metabolic pathways, including glycosylphosphatidylinositol-anchor biosynthesis, amino acid metabolism, sphingolipid metabolism, and fatty acid metabolism. iNOS deficiency during vascular endothelial dysfunction may affect the expression of TNFR-2, vascular adhesion factors, and the level of ROS via cellular metabolic changes, thereby attenuating vascular endothelial dysfunction.NEW & NOTEWORTHY Inducible nitric oxide synthase (iNOS) deficiency during vascular endothelial dysfunction may affect the expression of tumor necrosis factor receptor-2 and vascular adhesion factors via cellular metabolic changes, thereby attenuating vascular endothelial dysfunction.

Cellular Metabolomics Reveal the Mechanism Underlying the Anti-Atherosclerotic Effects of Aspirin Eugenol Ester on Vascular Endothelial Dysfunction

Aspirin eugenol ester (AEE) possesses anti-thrombotic, anti-atherosclerotic and anti-oxidative effects. The study aims to clarify the mechanism underlying the anti-atherosclerotic effects of AEE on vascular endothelial dysfunction. Both the high-fat diet (HFD)-induced atherosclerotic rat model and the H2O2-induced human umbilical vein endothelial cells (HUVECs) model were used to investigate the effects of AEE on vascular endothelial dysfunction. UPLC/QTOF-MS coupled with a multivariate data analysis method were used to profile the variations in the metabolites of HUVECs in response to different treatments. Pretreatment of HUVECs with AEE significantly ameliorated H2O2-induced apoptosis, the overexpression of E-selectin and VCAM-1, and the adhesion of THP-1 cells. Putative endogenous biomarkers associated with the inhibition of endothelial dysfunction were identified in HUVECs pretreated with AEE in the absence or presence of H2O2, and these biomarkers were involved in important metabolic pathways, including amino acid metabolism, carbohydrate metabolism, and glutathione metabolism. Moreover, in vivo, AEE also significantly reduced vascular endothelial dysfunction and decreased the overexpression of VCAM-1 and E-selectin. Based on our findings, the mechanism underlying the anti-atherosclerotic effects of AEE might be related to a reduction in vascular endothelial dysfunction mediated by ameliorating alterations in metabolism, inhibiting oxidative stress, and decreasing the expression of adhesion molecules.

Transcriptome and Metabolome Analyses in Exogenous FABP4- and FABP5-Treated Adipose-Derived Stem Cells

Adipose-derived stem cells (ADSC), which exist near adipocytes in adipose tissue, have been used as a potential tool of regenerative medicine. Lipid chaperones, fatty acid-binding protein 4 (FABP4) and 5 (FABP5), are abundantly expressed in adipocytes. FABP4 has recently been shown to be secreted from adipocytes during lipolysis in a non-classical pathway and may act as an adipokine. Here, we investigated the role of exogenous FABP4 and FABP5 in transcriptional and metabolic regulation in ADSC. FABP4 and FABP5 were little expressed in ADSC. However, both FABP4 and FABP5 were significantly induced after adipocyte differentiation of ADSC and were secreted from the differentiated adipocytes. Analysis of microarray data, including gene ontology enrichment analysis and cascade analysis of the protein-protein interaction network using a transcription factor binding site search, demonstrated that treatment of ADSC with FABP4 or FABP5 affected several kinds of genes related to inflammatory and metabolic responses and the process of cell differentiation. Notably, myogenic factors, including myocyte enhancer factors, myogenic differentiation 1 and myogenin, were modulated by treatment of ADSC with FABP4, indicating that exogenous FABP4 treatment is partially associated with myogenesis in ADSC. Metabolome analysis showed that treatment of ADSC with FABP4 and with FABP5 similarly, but differently in extent, promoted hydrolysis and/or uptake of lipids, consequentially together with enhancement of β oxidation, inhibition of downstream of the glycolysis pathway, accumulation of amino acids, reduction of nucleic acid components and increase in the ratio of reduced and oxidized nicotinamide adenine dinucleotide phosphates (NADPH/NADP+), an indicator of reducing power, and the ratio of adenosine triphosphate and adenosine monophosphate (ATP/AMP), an indicator of the energy state, in ADSC. In conclusion, secreted FABP4 and FABP5 from adipocytes as adipokines differentially affect transcriptional and metabolic regulation in ADSC near adipocytes. The adiposity condition in the host of regenerative medicine may affect characteristics of ADSC by exposure of the balance of FABP4 and FABP5.

Nutrient Deprivation Affects Salmonella Invasion and Its Interaction with the Gastrointestinal Microbiota

Salmonella enterica serovar Typhimurium (S. Typhimurium) is a foodborne enteric pathogen and a major cause of gastroenteritis in humans. It is known that molecules derived from the human fecal microbiota downregulate S. Typhimurium virulence gene expression and induce a starvation-like response. In this study, S. Typhimurium was cultured in minimal media to mimic starvation conditions such as that experienced by S. Typhimurium in the human intestinal tract, and the pathogen’s virulence in vitro and in vivo was measured. S. Typhimurium cultured in minimal media displayed a reduced ability to invade human epithelial cells in a manner that was at least partially independent of the Salmonella Pathogenicity Island 1 (SPI-1) type III secretion system. Nutrient deprivation did not, however, alter the ability of S. Typhimurium to replicate and survive inside epithelial cells. In a murine model of S. Typhimurium-induced gastroenteritis, prior cultivation in minimal media did not alter the pathogen’s ability to colonize mice, nor did it affect levels of gastrointestinal inflammation. Upon examining the post-infection fecal gastrointestinal microbiota, we found that specifically in the 129Sv/ImJ murine strain S. Typhimurium cultured in minimal media induced differential microbiota compositional shifts compared to that of S. Typhimurium cultured in rich media. Together these findings demonstrate that S. Typhimurium remains a potent pathogen even in the face of nutritional deprivation, but nevertheless that nutrient deprivation encountered in this environment elicits significant changes in the bacterium genetic programme, as well as its capacity to alter host microbiota composition.

Non-targeted metabolomic reveals the effect of salt stress on global metabolite of halotolerant yeast Candida versatilis and principal component analysis

As one of the major microbes in the soy sauce fermentation, Candida versatilis enriches the flavor and improves the quality of soy sauce. In this study, a combination of five different GC-MS and LC-MS-based metabolome analytical approaches was used to analyze the intracellular, extracellular and whole metabolites of C. versatilis. Our results found out that a total of 132, 244 and 267 different metabolites were detectable from the intracellular, extracellular and whole part, respectively. When exposed to 0. 9 and 18 % salt, respectively, 114, 123 and 129 different intracellular metabolites, 184, 200 and 178 extracellular metabolites and 177, 188 and 186 whole metabolites were detected, respectively. Our data showed that salt enhances the metabolic capacity of C. versatilis, especially its amino acid and enhances the synthesis and secretion of some metabolites of C. versatilis, especially the aldehydes and phenols, such as vanillin, guaiacol and 5-hydroxymethylfurfural. Our data also showed that special attention has to be paid to the generation of biogenic amines when C. versatilis was treated with salt.

Functional genomics of novel secondary metabolites from diverse cyanobacteria using untargeted metabolomics

Mass spectrometry-based metabolomics has become a powerful tool for the detection of metabolites in complex biological systems and for the identification of novel metabolites. We previously identified a number of unexpected metabolites in the cyanobacterium Synechococcus sp. PCC 7002, such as histidine betaine, its derivatives and several unusual oligosaccharides. To test for the presence of these compounds and to assess the diversity of small polar metabolites in other cyanobacteria, we profiled cell extracts of nine strains representing much of the morphological and evolutionary diversification of this phylum. Spectral features in raw metabolite profiles obtained by normal phase liquid chromatography coupled to mass spectrometry (MS) were manually curated so that chemical formulae of metabolites could be assigned. For putative identification, retention times and MS/MS spectra were cross-referenced with those of standards or available sprectral library records. Overall, we detected 264 distinct metabolites. These included indeed different betaines, oligosaccharides as well as additional unidentified metabolites with chemical formulae not present in databases of metabolism. Some of these metabolites were detected only in a single strain, but some were present in more than one. Genomic interrogation of the strains revealed that generally, presence of a given metabolite corresponded well with the presence of its biosynthetic genes, if known. Our results show the potential of combining metabolite profiling and genomics for the identification of novel biosynthetic genes.

Robust automated mass spectra interpretation and chemical formula calculation using mixed integer linear programming

Untargeted metabolite profiling using liquid chromatography and mass spectrometry coupled via electrospray ionization is a powerful tool for the discovery of novel natural products, metabolic capabilities, and biomarkers. However, the elucidation of the identities of uncharacterized metabolites from spectral features remains challenging. A critical step in the metabolite identification workflow is the assignment of redundant spectral features (adducts, fragments, multimers) and calculation of the underlying chemical formula. Inspection of the data by experts using computational tools solving partial problems (e.g., chemical formula calculation for individual ions) can be performed to disambiguate alternative solutions and provide reliable results. However, manual curation is tedious and not readily scalable or standardized. Here we describe an automated procedure for the robust automated mass spectra interpretation and chemical formula calculation using mixed integer linear programming optimization (RAMSI). Chemical rules among related ions are expressed as linear constraints and both the spectra interpretation and chemical formula calculation are performed in a single optimization step. This approach is unbiased in that it does not require predefined sets of neutral losses and positive and negative polarity spectra can be combined in a single optimization. The procedure was evaluated with 30 experimental mass spectra and was found to effectively identify the protonated or deprotonated molecule ([M + H](+) or [M - H](-)) while being robust to the presence of background ions. RAMSI provides a much-needed standardized tool for interpreting ions for subsequent identification in untargeted metabolomics workflows.

Metabolic footprinting of mutant libraries to map metabolite utilization to genotype

The discrepancy between the pace of sequencing and functional characterization of genomes is a major challenge in understanding complex microbial metabolic processes and metabolic interactions in the environment. Here, we identified and validated genes related to the utilization of specific metabolites in bacteria by profiling metabolite utilization in libraries of mutant strains. Untargeted mass spectrometry based metabolomics was used to identify metabolites utilized by Escherichia coli and Shewanella oneidensis MR-1. Targeted high-throughput metabolite profiling of spent media of 8042 individual mutant strains was performed to link utilization to specific genes. Using this approach we identified genes of known function as well as novel transport proteins and enzymes required for the utilization of tested metabolites. Specific examples include two subunits of a predicted ABC transporter encoded by the genes SO1043 and SO1044 required for the utilization of citrulline and a predicted histidase encoded by the gene SO3057 required for the utilization of ergothioneine by S. oneidensis. In vitro assays with purified proteins showed substrate specificity of SO3057 toward ergothioneine and histidine betaine in contrast to substrate specificity of a paralogous histidase SO0098 toward histidine. This generally applicable, high-throughput workflow has the potential both to discover novel metabolic capabilities of microorganisms and to identify the corresponding genes.

Merging multiple omics datasets in silico: statistical analyses and data interpretation

By the combinations of high-throughput analytical technologies in the fields of transcriptomics, proteomics, and metabolomics, we are now able to gain comprehensive and quantitative snapshots of the intracellular processes. Dynamic intracellular activities and their regulations can be elucidated by systematic observation of these multi-omics data. On the other hand, careful statistical analysis is necessary for such integration, since each of the omics layers as well as the specific analytical methodologies harbor different levels of noise and variations. Moreover, interpretation of such multitude of data requires an intuitive pathway context. Here we describe such statistical methods for the integration and comparison of multi-omics data, as well as the computational methods for pathway reconstruction, ID conversion, mapping, and visualization that play key roles for the efficient study of multi-omics information.

Substrate-baited nanoparticles: a catch and release strategy for enzyme recognition and harvesting

The isolation of a single type of protein from a complex mixture is vital for the characterization of the function, structure, and interactions of the protein of interest and is typically the most laborious aspect of the protein purification process. In this work, a model system is utilized to show the efficacy of synthesizing a "baited" nanoparticle to capture and recycle enzymes (proteins that catalyze chemical reactions) from crude cell lysate. Enzyme trapping and recycling is illustrated with the carbazole 1,9a-dioxygenase (CARDO) system, an enzyme important in bioremediation and natural product synthesis. The enzymes are baited with azide-modified carbazolyl moieties attached to poly(propargyl acrylate) nanoparticles through a click transformation. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis indicates the single-step procedure to immobilize the enzymes on the particles is capable of significantly concentrating the protein from raw lysate and sequestering all required components of the protein to maintain bioactivity. These results establish a universal model applicable to concentrating and extracting known substrate-protein pairs, but it can be an invaluable tool in recognizing unknown protein-ligand affinities.

Genomic analysis of the hydrocarbon-producing, cellulolytic, endophytic fungus Ascocoryne sarcoides

The microbial conversion of solid cellulosic biomass to liquid biofuels may provide a renewable energy source for transportation fuels. Endophytes represent a promising group of organisms, as they are a mostly untapped reservoir of metabolic diversity. They are often able to degrade cellulose, and they can produce an extraordinary diversity of metabolites. The filamentous fungal endophyte Ascocoryne sarcoides was shown to produce potential-biofuel metabolites when grown on a cellulose-based medium; however, the genetic pathways needed for this production are unknown and the lack of genetic tools makes traditional reverse genetics difficult. We present the genomic characterization of A. sarcoides and use transcriptomic and metabolomic data to describe the genes involved in cellulose degradation and to provide hypotheses for the biofuel production pathways. In total, almost 80 biosynthetic clusters were identified, including several previously found only in plants. Additionally, many transcriptionally active regions outside of genes showed condition-specific expression, offering more evidence for the role of long non-coding RNA in gene regulation. This is one of the highest quality fungal genomes and, to our knowledge, the only thoroughly annotated and transcriptionally profiled fungal endophyte genome currently available. The analyses and datasets contribute to the study of cellulose degradation and biofuel production and provide the genomic foundation for the study of a model endophyte system.

A renewable source of energy is a pressing global need. The biological conversion of lignocellulose to biofuels by microorganisms presents a promising avenue, but few organisms have been studied thoroughly enough to develop the genetic tools necessary for rigorous experimentation. The filamentous-fungal endophyte A. sarcoides produces metabolites when grown on a cellulose-based medium that include eight-carbon volatile organic compounds, which are potential biofuel targets. Here we use broadly applicable methods including genomics, transcriptomics, and metabolomics to explore the biofuel production of A. sarcoides. These data were used to assemble the genome into 16 scaffolds, to thoroughly annotate the cellulose-degradation machinery, and to make predictions for the production pathway for the eight-carbon volatiles. Extremely high expression of the gene swollenin when grown on cellulose highlights the importance of accessory proteins in addition to the enzymes that catalyze the breakdown of the polymers. Correlation of the production of the eight-carbon biofuel-like metabolites with the expression of lipoxygenase pathway genes suggests the catabolism of linoleic acid as the mechanism of eight-carbon compound production. This is the first fungal genome to be sequenced in the family Helotiaceae, and A. sarcoides was isolated as an endophyte, making this work also potentially useful in fungal systematics and the study of plant–fungus relationships.

Remodeling of hepatic metabolism and hyperaminoacidemia in mice deficient in proglucagon-derived peptides

Glucagon is believed to be one of the most important peptides for upregulating blood glucose levels. However, homozygous glucagon-green fluorescent protein (gfp) knock-in mice (Gcg(gfp/gfp): GCGKO) are normoglycemic despite the absence of proglucagon-derived peptides, including glucagon. To characterize metabolism in the GCGKO mice, we analyzed gene expression and metabolome in the liver. The expression of genes encoding rate-limiting enzymes for gluconeogenesis was only marginally altered. On the other hand, genes encoding enzymes involved in conversion of amino acids to metabolites available for the tricarboxylic acid cycle and/or gluconeogenesis showed lower expression in the GCGKO liver. The expression of genes involved in the metabolism of fatty acids and nicotinamide was also altered. Concentrations of the metabolites in the GCGKO liver were altered in manners concordant with alteration in the gene expression patterns, and the plasma concentrations of amino acids were elevated in the GCGKO mice. The insulin concentration in serum and phosphorylation of Akt protein kinase in liver were reduced in GCGKO mice. These results indicated that proglucagon-derived peptides should play important roles in regulating various metabolic pathways, especially that of amino acids. Serum insulin concentration is lowered to compensate the impacts of absent proglucagon-derived peptide on glucose metabolism. On the other hand, impacts on other metabolic pathways are only partially compensated by reduced insulin action.

Metabolomic analysis of trypanosomatid protozoa

Metabolomics aims to measure all low molecular weight chemicals within a given system in a manner analogous to transcriptomics, proteomics and genomics. In this review we highlight metabolomics approaches that are currently being applied to the kinetoplastid parasites, Trypanosoma brucei and Leishmania spp. The use of untargeted metabolomics approaches, made possible through advances in mass spectrometry and informatics, and stable isotope labelling has increased our understanding of the metabolism in these organisms beyond the views established using classical biochemical approaches. Set within the context of metabolic networks, predicted using genome-wide reconstructions of metabolism, new hypotheses on how to target aspects of metabolism to design new drugs against these protozoa are emerging.

Prediction of metabolic reactions based on atomic and molecular properties of small-molecule compounds

MOTIVATION

Our knowledge of the metabolites in cells and their reactions is far from complete as revealed by metabolomic measurements that detect many more small molecules than are documented in metabolic databases. Here, we develop an approach for predicting the reactivity of small-molecule metabolites in enzyme-catalyzed reactions that combines expert knowledge, computational chemistry and machine learning.

RESULTS

We classified 4843 reactions documented in the KEGG database, from all six Enzyme Commission classes (EC 1-6), into 80 reaction classes, each of which is marked by a characteristic functional group transformation. Reaction centers and surrounding local structures in substrates and products of these reactions were represented using SMARTS. We found that each of the SMARTS-defined chemical substructures is widely distributed among metabolites, but only a fraction of the functional groups in these substructures are reactive. Using atomic properties of atoms in a putative reaction center and molecular properties as features, we trained support vector machine (SVM) classifiers to discriminate between functional groups that are reactive and non-reactive. Classifier accuracy was assessed by cross-validation analysis. A typical sensitivity [TP/(TP+FN)] or specificity [TN/(TN+FP)] is ≈0.8. Our results suggest that metabolic reactivity of small-molecule compounds can be predicted with reasonable accuracy based on the presence of a potentially reactive functional group and the chemical features of its local environment.

AVAILABILITY

The classifiers presented here can be used to predict reactions via a web site (http://cellsignaling.lanl.gov/Reactivity/). The web site is freely available.