Metabolomics atlas of oral 13C-glucose tolerance test in mice

Changes in Isoleucine, Sarcosine, and Dimethylglycine During OGTT as Risk Factors for Diabetes

CONTEXT

Current metabolomics studies in diabetes have focused on the fasting state, while only a few have addressed the satiated state.

OBJECTIVE

We combined the oral glucose tolerance test (OGTT) and metabolomics to examine metabolite-level changes in populations with different glucose tolerance statuses and to evaluate the potential risk of these changes for diabetes.

METHODS

We grouped participants into those with normal glucose tolerance (NGT), impaired glucose regulation (IGR), and newly diagnosed type 2 diabetes (NDM). During the OGTT, serum was collected at 0, 30, 60, 120, and 180 minutes. We evaluated the changes in metabolite levels during the OGTT and compared metabolic profiles among the 3 groups. The relationship between metabolite levels during the OGTT and risk of diabetes and prediabetes was analyzed using a generalized estimating equation (GEE). The regression results were adjusted for sex, body mass index, fasting insulin levels, heart rate, smoking status, and blood pressure.

RESULTS

Glucose intake altered metabolic profile and induced an increase in glycolytic intermediates and a decrease in amino acids, glycerol, ketone bodies, and triglycerides. Isoleucine levels differed between the NGT and NDM groups and between the NGT and IGR groups. Changes in sarcosine levels during the OGTT in the diabetes groups were opposite to those in glycine levels. GEE analysis revealed that during OGTT, isoleucine, sarcosine, and acetic acid levels were associated with NDM risks, and isoleucine and acetate levels with IGR risks.

CONCLUSION

Metabolic profiles differ after glucose induction in individuals with different glucose tolerance statuses. Changes in metabolite levels during OGTT are potential risk factors for diabetes development.

Lipidome isotope labelling of gut microbes (LILGM): A method of discovering endogenous microbial lipids

Lipidomics analysis of gut microbiome has become critical in recent surge of extensive human disease studies that investigate microbiome contributions. However, challenges remain in comprehending the origins of thousands of lipid species produced by the diverse microbes. Here, we proposed the development and utilization of a liquid chromatography-mass spectrometry-based approach, named lipidome isotope labelling of gut microbes (LILGM), which enables confident detection and identification of endogenous gut microbial lipidome via 13C/15N labeling strategy and high-resolution mass spectrometry. Our method leveraged in vitro microbial cultures and stable isotope-labeled 13C and 15N, allowing a reasonable degree of isotope incorporation into microbial lipids over short-term of inoculation. We then systematically detected the mass spectral patterns of 182 labeled lipid species by our in-house data analysis pipeline. Further bioinformatics analyses confidently identified biologically relevant microbial lipids from lipid classes such as diacylglycerols (DGs), fatty acids (FAs), phosphatidylglycerols (PGs), and phosphatidylethanolamines (PEs) that may have profound impacts to human physiology. Our study also demonstrated the application of LILGM by showcasing the confident detection of dysregulated microbial lipids post antibiotic perturbation. The debiased sparse partial correlation analysis provides insights into lipid metabolism intricacies. Overall, our method can provide unambiguous analyses to the endogenous microbial lipids in given biological context, and can also instantly reflect the lipidomic changes of gut microbes in response to environmental factors. We believe our LILGM approach has the potential to provide new body of knowledge by combining promising analytical approaches for sensitive and specific lipid detection to support functional microbiome studies.

Tracers and Imaging of Fatty Acid and Energy Metabolism of Human Adipose Tissues

White adipose tissue and brown adipose tissue (WAT and BAT) regulate fatty acid metabolism and control lipid fluxes to other organs. Dysfunction of these key metabolic processes contributes to organ insulin resistance and inflammation leading to chronic diseases such as type 2 diabetes, metabolic dysfunction-associated steatohepatitis, and cardiovascular diseases. Metabolic tracers combined with molecular imaging methods are powerful tools for the investigation of these pathogenic mechanisms. Herein, I review some of the positron emission tomography and magnetic resonance imaging methods combined with stable isotopic metabolic tracers to investigate fatty acid and energy metabolism, focusing on human WAT and BAT metabolism. I will discuss the complementary strengths offered by these methods for human investigations and current gaps in the field.

Using Data-Dependent and -Independent Hybrid Acquisitions for Fast Liquid Chromatography-Based Untargeted Lipidomics

Untargeted lipidomics using liquid chromatography (LC) coupled with tandem mass spectrometry (MS) is essential for large cohort studies. Using a fast LC gradient of less than 10 min for the rapid screening of lipids decreases the annotation rate, because of the lower coverage of the MS/MS spectra caused by the narrow peak width. A systematic procedure is proposed in this study to achieve a high annotation rate in fast LC-based untargeted lipidomics by integrating data-dependent acquisition (DDA) and sequential window acquisition of all-theoretical mass spectrometry data-independent acquisition (SWATH-DIA) techniques using the updated MS-DIAL program. This strategy uses variable SWATH-DIA methods for quality control (QC) samples, which are a mixture of biological samples that were analyzed multiple times to correct the MS signal drift. In contrast, biological samples are analyzed using DDA to facilitate the structural elucidation of lipids using the pure spectrum to the maximum extent. The workflow is demonstrated using an 8.6 min LC gradient, where the QC samples are analyzed using five different SWATH-DIA methods. The use of both DDA and SWATH-DIA achieves a 1.7-fold annotation coverage from publicly available benchmark data obtained using a fast LC-DDA-MS technique and offers 95.3% lipid coverage, as compared to the benchmark data set from a 25 min LC gradient. This study demonstrates that harmonized improvements in analytical conditions and informatics tools provide a comprehensive lipidome in fast LC-based untargeted lipidomics, not only for large-scale studies but also for small-scale experiments, contributing to both clinical applications and basic biology.

Two mitochondrial DNA polymorphisms modulate cardiolipin binding and lead to synthetic lethality

Genetic screens have been used extensively to probe interactions between nuclear genes and their impact on phenotypes. Probing interactions between mitochondrial genes and their phenotypic outcome, however, has not been possible due to a lack of tools to map the responsible polymorphisms. Here, using a toolkit we previously established in Drosophila, we isolate over 300 recombinant mitochondrial genomes and map a naturally occurring polymorphism at the cytochrome c oxidase III residue 109 (CoIII109) that fully rescues the lethality and other defects associated with a point mutation in cytochrome c oxidase I (CoIT300I). Through lipidomics profiling, biochemical assays and phenotypic analyses, we show that the CoIII109 polymorphism modulates cardiolipin binding to prevent complex IV instability caused by the CoIT300I mutation. This study demonstrates the feasibility of genetic interaction screens in animal mitochondrial DNA. It unwraps the complex intra-genomic interplays underlying disorders linked to mitochondrial DNA and how they influence disease expression.

Jean DJ, Tieu T, Ladd B, Hilbert B, Wang W, Alltucker JT, Manimala S, Kryukov GV, Brooijmans N, Dowdell G, Jonsson P, Huff M, Guzman-Perez A, Jackson EL, Goncalves MD, Stuart DD. STX-478, a Mutant-Selective, Allosteric PI3Kα Inhibitor Spares Metabolic Dysfunction and Improves Therapeutic Response in PI3Kα-Mutant Xenografts

Phosphoinositide 3-kinase α (PIK3CA) is one of the most mutated genes across cancers, especially breast, gynecologic, and head and neck squamous cell carcinoma tumors. Mutations occur throughout the gene, but hotspot mutations in the helical and kinase domains predominate. The therapeutic benefit of isoform-selective PI3Kα inhibition was established with alpelisib, which displays equipotent activity against the wild-type and mutant enzyme. Inhibition of wild-type PI3Kα is associated with severe hyperglycemia and rash, which limits alpelisib use and suggests that selectively targeting mutant PI3Kα could reduce toxicity and improve efficacy. Here we describe STX-478, an allosteric PI3Kα inhibitor that selectively targets prevalent PI3Kα helical- and kinase-domain mutant tumors. STX-478 demonstrated robust efficacy in human tumor xenografts without causing the metabolic dysfunction observed with alpelisib. Combining STX-478 with fulvestrant and/or cyclin-dependent kinase 4/6 inhibitors was well tolerated and provided robust and durable tumor regression in ER+HER2- xenograft tumor models.

SIGNIFICANCE

These preclinical data demonstrate that the mutant-selective, allosteric PI3Kα inhibitor STX-478 provides robust efficacy while avoiding the metabolic dysfunction associated with the nonselective inhibitor alpelisib. Our results support the ongoing clinical evaluation of STX-478 in PI3Kα-mutated cancers, which is expected to expand the therapeutic window and mitigate counterregulatory insulin release. See related commentary by Kearney and Vasan, p. 2313. This article is featured in Selected Articles from This Issue, p. 2293.

Metabolomics and lipidomics strategies in modern drug discovery and development

Metabolomics and lipidomics have an increasingly pivotal role in drug discovery and development. In the context of drug discovery, monitoring changes in the levels or composition of metabolites and lipids relative to genetic variations yields functional insights, bolstering human genetics and (meta)genomic methodologies. This approach also sheds light on potential novel targets for therapeutic intervention. In the context of drug development, metabolite and lipid biomarkers contribute to enhanced success rates, promising a transformative impact on precision medicine. In this review, we deviate from analytical chemist-focused perspectives, offering an overview tailored to drug discovery. We provide introductory insight into state-of-the-art mass spectrometry (MS)-based metabolomics and lipidomics techniques utilized in drug discovery and development, drawing from the collective expertise of our research teams. We comprehensively outline the application of metabolomics and lipidomics in advancing drug discovery and development, spanning fundamental research, target identification, mechanisms of action, and the exploration of biomarkers.

Stachydrine, N-acetylornithine and trimethylamine N-oxide levels as candidate milk biomarkers of maternal consumption of an obesogenic diet during lactation

We aimed to evaluate whether improving maternal diet during lactation in diet-induced obese rats reverts the impact of western diet (WD) consumption on the metabolome of milk and offspring plasma, as well as to identify potential biomarkers of these conditions. Three groups of dams were followed: control-dams (CON-dams), fed with standard diet (SD); WD-dams, fed with WD prior and during gestation and lactation; and reversion-dams (REV-dams), fed as WD-dams but moved to SD during lactation. Metabolomic analysis was performed in milk at lactation days 5, 10, and 15, and in plasma from their male and female offspring at postnatal day 15. Milk of WD-dams presented, throughout lactation and compared to CON-dams, altered profiles of amino acids and of the carnitine pool, accompanied by changes in other polar metabolites, being stachydrine, N-acetylornithine, and trimethylamine N-oxide the most relevant and discriminatory metabolites between groups. The plasma metabolome profile was also altered in the offspring of WD-dams in a sex-dependent manner, and stachydrine, ergothioneine and the acylcarnitine C12:1 appeared as the top three most discriminating metabolites in both sexes. Metabolomic changes were largely normalized to control levels both in the milk of REV-dams and in the plasma of their offspring. We have identified a set of polar metabolites in maternal milk and in the plasma of the offspring whose alterations may indicate maternal intake of an unbalanced diet during gestation and lactation. Levels of these metabolites may also reflect the beneficial effects of implementing a healthier diet during lactation.

LORA, Lipid Over-Representation Analysis Based on Structural Information

With the increasing number of lipidomic studies, there is a need for an efficient and automated analysis of lipidomic data. One of the challenges faced by most existing approaches to lipidomic data analysis is lipid nomenclature. The systematic nomenclature of lipids contains all available information about the molecule, including its hierarchical representation, which can be used for statistical evaluation. The Lipid Over-Representation Analysis (LORA) web application (https://lora.metabolomics.fgu.cas.cz) analyzes this information using the Java-based Goslin framework, which translates lipid names into a standardized nomenclature. Goslin provides the level of lipid hierarchy, including information on headgroups, acyl chains, and their modifications, up to the "complete structure" level. LORA allows the user to upload the experimental query and reference data sets, select a grammar for lipid name normalization, and then process the data. The user can then interactively explore the results and perform lipid over-representation analysis based on selected criteria. The results are graphically visualized according to the lipidome hierarchy. The lipids present in the most over-represented terms (lipids with the highest number of enriched shared structural features) are defined as Very Important Lipids (VILs). For example, the main result of a demo data set is the information that the query is significantly enriched with "glycerophospholipids" containing "acyl 20:4" at the "sn-2 position". These terms define a set of VILs (e.g., PC 18:2/20:4;O and PE 16:0/20:4(5,8,10,14);OH). All results, graphs, and visualizations are summarized in a report. LORA is a tool focused on the smart mining of epilipidomics data sets to facilitate their interpretation at the molecular level.

Combining albumin deficiency and acute exercise reduces hepatic lipid droplet size in mice

Hepatic lipid droplets (LDs) are implicated in ectopic lipid accumulation. The core of LDs, triacylglycerol (TAG), is synthesized from the esterification of fatty acids to a glycerol-3-phosphate (G-3-P) backbone. Albumin transports plasma free fatty acids, and previously albumin knockout (Alb^-/-) mice were shown to exhibit lower hepatic TAG levels than wildtype (WT). Exercise is a beneficial strategy to alter hepatic metabolism, but its impacts on reducing hepatic lipids are far from satisfactory. The aim of this study was to investigate the combined effect of albumin deficiency and acute exercise on hepatic LDs. Eight-week-old male Alb^-/- and WT mice were divided into sedentary and exercise groups. Exercised mice performed a 30-min high-intensity exercise bout. Results showed that sedentary Alb^-/- mice had smaller hepatic LDs (P < 0.0001), associated with mitochondria, while WT mice exhibited larger LDs, surrounded by glycogen granules. Following acute exercise, hepatic LDs in Alb^-/- mice reduced by 40% in size, while in WT increased by 14% (P < 0.0001). The maintenance of WT hepatic LDs was associated with elevated G-3-P level (P < 0.05), potentially derived from glycogen (R = -0.32, %change in glycogen versus LD content, P < 0.05). The reduction in Alb^-/- mice LDs after exercise was possibly due to their low glycogen level. In conclusion, Alb^-/- mice exhibited an enhanced capacity for reducing hepatic LD size and content in response to exercise. These findings suggest that modulating albumin's functions combined with exercise could be a potential strategy to reduce ectopic lipid deposition in the liver.

Short-Term Stability of Serum and Liver Extracts for Untargeted Metabolomics and Lipidomics

Thermal reactions can significantly alter the metabolomic and lipidomic content of biofluids and tissues during storage. In this study, we investigated the stability of polar metabolites and complex lipids in dry human serum and mouse liver extracts over a three-day period under various temperature conditions. Specifically, we tested temperatures of -80 °C (freezer), -24 °C (freezer), -0.5 °C (polystyrene box with gel-based ice packs), +5 °C (refrigerator), +23 °C (laboratory, room temperature), and +30 °C (thermostat) to simulate the time between sample extraction and analysis, shipping dry extracts to different labs as an alternative to dry ice, and document the impact of higher temperatures on sample integrity. The extracts were analyzed using five fast liquid chromatography-mass spectrometry (LC-MS) methods to screen polar metabolites and complex lipids, and over 600 metabolites were annotated in serum and liver extracts. We found that storing dry extracts at -24 °C and partially at -0.5 °C provided comparable results to -80 °C (reference condition). However, increasing the storage temperatures led to significant changes in oxidized triacylglycerols, phospholipids, and fatty acids within three days. Polar metabolites were mainly affected at storage temperatures of +23 °C and +30 °C.

Optimization of Mobile Phase Modifiers for Fast LC-MS-Based Untargeted Metabolomics and Lipidomics

Liquid chromatography-mass spectrometry (LC-MS) is the method of choice for the untargeted profiling of biological samples. A multiplatform LC-MS-based approach is needed to screen polar metabolites and lipids comprehensively. Different mobile phase modifiers were tested to improve the electrospray ionization process during metabolomic and lipidomic profiling. For polar metabolites, hydrophilic interaction LC using a mobile phase with 10 mM ammonium formate/0.125% formic acid provided the best performance for amino acids, biogenic amines, sugars, nucleotides, acylcarnitines, and sugar phosphate, while reversed-phase LC (RPLC) with 0.1% formic acid outperformed for organic acids. For lipids, RPLC using a mobile phase with 10 mM ammonium formate or 10 mM ammonium formate with 0.1% formic acid permitted the high signal intensity of various lipid classes ionized in ESI(+) and robust retention times. For ESI(-), the mobile phase with 10 mM ammonium acetate with 0.1% acetic acid represented a reasonable compromise regarding the signal intensity of the detected lipids and the stability of retention times compared to 10 mM ammonium acetate alone or 0.02% acetic acid. Collectively, we show that untargeted methods should be evaluated not only on the total number of features but also based on common metabolites detected by a specific platform along with the long-term stability of retention times.

The role of peroxiredoxin 6 in biosynthesis of FAHFAs

Peroxiredoxin 6 (Prdx6) is a multifunctional enzyme, a unique member of the peroxiredoxin family, with an important role in antioxidant defense. Moreover, it has also been linked with the biosynthesis of anti-inflammatory and anti-diabetic lipids called fatty acid esters of hydroxy fatty acids (FAHFAs) and many diseases, including cancer, inflammation, and metabolic disorders. Here, we performed metabolomic and lipidomic profiling of subcutaneous adipose tissue from mouse models with genetically modified Prdx6. Deletion of Prdx6 resulted in reduced levels of FAHFAs containing 13-hydroxylinoleic acid (13-HLA). Mutation of Prdx6 C47S impaired the glutathione peroxidase activity and reduced FAHFA levels, while D140A mutation, responsible for phospholipase A2 activity, showed only minor effects. Targeted analysis of oxidized phospholipids and triacylglycerols in adipocytes highlighted a correlation between FAHFA and hydroxy fatty acid production by Prdx6 or glutathione peroxidase 4. FAHFA regioisomer abundance was negatively affected by the Prdx6 deletion, and this effect was more pronounced in longer and more unsaturated FAHFAs. The predicted protein model of Prdx6 suggested that the monomer-dimer transition mechanism might be involved in the repair of longer-chain peroxidized phospholipids bound over two monomers and that the role of Prdx6 in FAHFA synthesis might be restricted to branching positions further from carbon 9. In conclusion, our work linked the peroxidase activity of Prdx6 with the levels of FAHFAs in adipose tissue.

Exogenous detection of 13C-glucose metabolism in tumor and diet-induced obesity models

Metabolic rewiring is a hallmark feature prevalent in cancer cells as well as insulin resistance (IR) associated with diet-induced obesity (DIO). For instance, tumor metabolism shifts towards an enhanced glycolytic state even under aerobic conditions. In contrast, DIO triggers lipid-induced IR by impairing insulin signaling and reducing insulin-stimulated glucose uptake. Based on physiological differences in systemic metabolism, we used a breath analysis approach to discriminate between different pathological states using glucose oxidation as a readout. We assessed glucose utilization in lung cancer-induced cachexia and DIO mouse models using a U-13C glucose tracer and stable isotope sensors integrated into an indirect calorimetry system. Our data showed increased 13CO2 expired by tumor-bearing (TB) mice and a reduction in exhaled 13CO2 in the DIO model. Taken together, our findings illustrate high glucose uptake and consumption in TB animals and decreased glucose uptake and oxidation in obese mice with an IR phenotype. Our work has important translational implications for the utility of stable isotopes in breath-based detection of glucose homeostasis in models of lung cancer progression and DIO.

In vivo transomic analyses of glucose-responsive metabolism in skeletal muscle reveal core differences between the healthy and obese states

Metabolic regulation in skeletal muscle is essential for blood glucose homeostasis. Obesity causes insulin resistance in skeletal muscle, leading to hyperglycemia and type 2 diabetes. In this study, we performed multiomic analysis of the skeletal muscle of wild-type (WT) and leptin-deficient obese (ob/ob) mice, and constructed regulatory transomic networks for metabolism after oral glucose administration. Our network revealed that metabolic regulation by glucose-responsive metabolites had a major effect on WT mice, especially carbohydrate metabolic pathways. By contrast, in ob/ob mice, much of the metabolic regulation by glucose-responsive metabolites was lost and metabolic regulation by glucose-responsive genes was largely increased, especially in carbohydrate and lipid metabolic pathways. We present some characteristic metabolic regulatory pathways found in central carbon, branched amino acids, and ketone body metabolism. Our transomic analysis will provide insights into how skeletal muscle responds to changes in blood glucose and how it fails to respond in obesity.

Triacylglycerols containing branched palmitic acid ester of hydroxystearic acid (PAHSA) are present in the breast milk and hydrolyzed by carboxyl ester lipase

Breast milk is a complex mixture containing underexplored bioactive lipids. We performed an observational case-control study to compare the impact of delivery mode: caesarean section (CS) and vaginal birth (VB); and term (preterm and term delivery) on the levels of lipokines in human milk at different stages of lactation. Metabolomic analysis of the milk identified triacylglycerol estolides as a metabolic reservoir of the anti-inflammatory lipid mediator 5-palmitic acid ester of hydroxystearic acid (5-PAHSA). We found that triacylglycerol estolides were substrates of carboxyl ester lipase and 5-PAHSA-containing lipids were the least preferred substrates among tested triacylglycerol estolide isomers. This explained exceptionally high colostrum levels of 5-PAHSA in the VB group. CS and preterm birth negatively affected colostrum lipidome, including 5-PAHSA levels, but the lipidomic profiles normalized in mature milk. Mothers delivering term babies vaginally produce colostrum rich in 5-PAHSA, which could contribute to the prevention of intestinal inflammation in newborns.