Metabolomics and lipidomics strategies in modern drug discovery and development

Abstract

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 enhance 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.

Teaser:

Here, we highlight the pivotal roles of metabolomics and lipidomics in drug discovery and development and explore genetic links, functional insights, novel targets, and promising biomarkers driving precision medicine forward.

Introduction

Metabolites have a crucial role in cellular energy production by providing substrates for glycolysis/fermentation and oxidative phosphorylation. Additionally, they act as building blocks for large biomolecules and cell membranes, as well as signaling messengers. A biological sample can contain thousands of metabolites, covering a diverse range of biochemical classes, such as amino acids, biogenic amines, organic acids, nucleotides, and lipids.^1-3

Alterations in metabolite levels are intricately tied to diverse diseases, revealing insights into disease etiology, status, therapeutic response, and biological process changes.⁴ The systematic study of metabolites found within biological systems is termed ‘metabolomics’.^5,6 Whereas metabolomics aims to identify and quantify the composition of metabolites, lipidomics is a subset of metabolomics that focuses specifically on lipid metabolites. Throughout this review, we use the term ‘metabolomics’ to encompass both metabolomic and lipidomic investigations.^7-9

The profile of metabolites, or the ‘metabolome’, is determined by two factors: the biological activity of cells, which reflects the genomic variation in biological systems (e.g., individuals), and the impact of environmental factors, such as drugs, diet, lifestyle, and the microbiome on those biological systems (Figure 1). By capturing the intricate and multifaceted landscape of the metabolome and tracking changes in the levels or composition of metabolites, metabolomics can depict distinct molecular phenotypes indicative of the physiological and pathological states of an organism (Figure 1).^5,6

Figure 1.

The intricate network of processes that contribute to the human metabolome highlights the nonlinear and multifactorial nature of the genotype–phenotype relationship. Although we can measure genes, we cannot always differentiate between their active and inactive forms. Moreover, transcription of DNA into RNA and translation into proteins is controlled by a series of biological processes, rendering transcript levels insufficient to predict protein levels and explain genotype–phenotype relationships.¹⁵⁹ In fact, the correlation between gene expression and protein measurement is relatively poor.^159,160 Proteins are further regulated by protein–protein interactions, post-translational modifications (PTMs), and feedback inhibition and activation mechanisms.^161,162 Microbiome, environmental, and lifestyle factors contribute to the complexity of the human metabolome directly by regulating cellular processes and indirectly by influencing the influx of metabolites. Ultimately, all these factors converge to define an individual's metabolome.

The idea of using metabolites to evaluate health and disease is not a new concept. Ancient medical practices utilized sight, taste, or smell to assess metabolites from human samples. However, in modern times, metabolites have revolutionized the field of medicine by becoming the foundation for clinical diagnostic tests and medical treatments. In fact, >95% of clinical diagnostic tests measure specific metabolites, and ~90% of drugs are small molecules.^10,11 As such, metabolites have become an essential element of modern medicine, and metabolite assays are now considered routine tests for diagnosing and monitoring various diseases throughout an individual's lifespan. For example, at birth, clinicians measure a comprehensive panel of blood metabolites to diagnose inborn errors of metabolism (see Glossary).¹² blood levels of creatinine are assessed to diagnose renal dysfunction¹³ and measurements of lipid metabolites, such as cholesterol and triglycerides, form the basis of cardiovascular screening tests.¹⁴ More recently, a class of lipids known as ceramides, found in easily accessible body fluids, has emerged as promising biomarker candidates. A ceramide panel is already commercially available to select clinical laboratories, aiding in predicting adverse cardiovascular events in patients with advancing coronary artery disease.¹⁵

The remarkable advances in the sensitivity and specificity of MS and nuclear magnetic resonance (NMR) have led to the development of sophisticated analytical tools that can precisely measure an ever-increasing number of metabolites in biological samples (Box 1).¹⁶ Compared with only a few decades ago, we can now measure metabolites in biological samples hundreds of times faster and with greater specificity. For instance, in addition to measuring total triglyceride levels in the blood, we can now characterize their chemical structure and fatty acid composition in exquisite detail. This presents an unprecedented opportunity to accurately and comprehensively describe molecular phenotypes, thereby enhancing the potential for discovering novel targets for therapeutic interventions and valuable biomarkers.^2,17-19

Box 1. MS and NMR.

MS and NMR are the most used analytical techniques in metabolomics analysis. NMR has emerged as a mature and robust technology: quantitative, nondestructive, and capable of high-throughput performance.^163,164 Nevertheless, NMR is not sensitive enough to measure low-abundance metabolites; thus, quantitation of individual metabolites is often difficult, owing to signal overlap. Consequently, in fast-paced drug discovery and development environment, MS has become the workhorse of metabolomic analysis primarily because of its sensitivity, which allows for analyzing metabolites present at low abundance or in small sample volumes, and its quick turnaround time.

In this review, we present a preliminary overview of advanced MS approaches used in metabolomics, drawing from the collective expertise of our research teams. Our aim here is not to provide a review primarily focused on analytical chemistry aspects, for which we direct readers to specific reviews. Instead, we aim to provide a tailored overview within the scope of drug discovery. We explore the diverse applications of metabolomics across various stages of drug discovery and development, encompassing foundational research, target identification, mechanism-of-action studies, and the creation of clinically applicable biomarkers. Our endeavor also involves incorporating recent industry-led studies highlighting the use of metabolomics and lipidomics in drug discovery and development. Ultimately, our goal is to inspire and offer guidance to multidisciplinary scientists operating within collaborative environments, such as matrixed organizations and research consortia, by shedding light on the potential applications of metabolomics in their research endeavors.

Stages to approach metabolomics studies

Generally, a metabolomic study involves three stages: pre-analytical, analytical, and post-analytical (Figure 2). These stages encompass all aspects of the study, from sample collection and preparation to data analysis and interpretation. Itis crucial to consider these stages diligently, and treat them as interdependent to align them with the ultimate objectives of the project.

Figure 2.

Typical workflow for a metabolomic study in a collaborative environment. Metabolomics contributes to a data-driven, in-depth description of the molecular phenotypes allowing for the identification of molecular subtypes that could lead to the identification of common altered biochemical pathways across cellular, animal, and human experiments. Metabolite data sets from metabolomics studies can be integrated into broader investigations that encompass additional ‘omics data and other molecular information. This integrated database holds the potential to unveil disease etiology, status, therapeutic response, and shifts in biological processes.

Pre-analytical approaches

Pre-analytical steps are crucial to ensure that the metabolomic measurements accurately reflect the endogenous levels and composition of metabolites in the sample. Factors, such as sample collection, handling, and storage, can affect metabolite stability and introduce variability in the measurements. For example, enzymatic activity can continue in biological samples after collection, leading to changes in metabolite levels. Exposure to oxygen, light, or other environmental factors can also cause metabolite degradation or modification. Therefore, careful attention must be paid to pre-analytical steps to ensure that the samples are collected, processed, and stored in a standardized and controlled manner.^4,20-22

To ensure accurate and reliable results in human metabolomics studies, it is crucial to consider the study design and carefully select participants based on variables that might affect metabolite levels in biological specimens. These variables include age, sex, body-mass index, ethnicity, and current medications or dietary supplement intake, among others. When analyzing postmortem tissue samples, the interval between death and specimen collection should also be considered. Moreover, when analyzing specific tissue regions or subregions, meticulous anatomical selection is paramount. In specialized analytical approaches, such as fluxomics or spatial metabolomics, strategic decisions regarding labeled nutrients or embedding techniques are crucial. Overlooking these factors, which have a pivotal role in metabolomics data analysis and interpretation, can introduce confounding elements that hinder the effectiveness and influence of the study.

To ensure the accuracy and reproducibility of metabolomics studies, it is crucial to standardize the sample collection and handling procedures, including the time of day for collection, fasting status, collection tubes, centrifugation steps before freezing the samples, number of freeze–thaw cycles, consumables used, and sample storage conditions. Standardized operating procedures (SOPs) should be established and followed consistently to minimize variability in sample handling and ensure reproducibility of results. All biological samples should be handled as consistently as possible, and samples should be randomized during collection and processing. Failure to follow these guidelines can negatively impact the robustness and reproducibility of the data, making data analysis and interpretation more complicated^4,20-22

Analytical approaches

Once pre-analytical steps have been considered, it is important to standardize the procedures for sample preparation and analysis when designing a metabolomics study. Adhering to standardized procedures is crucial to ensure the reproducibility of results and enable interlaboratory comparison of data.

In metabolomics, the initial step is sample preparation, which involves extracting metabolites from proteins and other constituents present in the biological sample matrix. It is crucial to detail every step in a SOP to minimize variability and standardize the process across multiple operators. To further reduce human error, it is advisable to automate procedures for sample preparation when possible.

Next, the complex mixture of metabolites is separated using chromatography or ion mobility or is analyzed directly by MS. The sensitivity and selectivity of MS analysis enable the measurement of numerous and diverse metabolites in biological samples. For a comprehensive overview of chromatographic, ion mobility and MS methods applicable to metabolomics and lipidomics analyses, we direct readers to ^23,24.

The final step in designing a metabolomics study involves selecting analytical approaches appropriate for answering the scientific question(s) of interest. Various approaches can be adopted, including discovery metabolomics, biomarker validation, fluxomics, and spatial metabolomics.

Discovery metabolomics

Discovery metabolomics is an approach to screening as many metabolites as possible in a biological sample without bias. This approach generates novel hypotheses about the potential role of specific metabolites or metabolic pathways in disease and drug mechanisms and their possible use as biomarkers and therapeutic targets. To compare metabolic profiles between different groups, such as healthy versus diseased, control versus treated, wild-type versus genetically modified, or responders versus non-responders, various statistical tools can be used. The data obtained from this approach could also help characterize the heterogeneity of a patient population by defining individuals' metabotypes, leading to improved patient selection and precision medicine.

The unbiased, discovery metabolomics approach can be particularly useful when investigating the molecular signature of complex, multifactorial diseases. These types of disease have multiple underlying causes, resulting in a heterogeneous patient population. In a recent study, for example, the plasma metabolome contributed to the molecular characterization of five different subtypes of type 2 diabetes mellitus (T2DM) in three independent cohorts comprising 15 940 individuals. This approach provided novel biological insights into the diverse etiological processes that would not be evident when T2DM is viewed as a homogeneous disease.²⁵

A discovery metabolomics approach can also be useful in identifying biomarkers linked to mutations in single genes. For example, a mutation in the gene encoding progranulin (GRN) can lead to frontotemporal dementia resulting from a deficiency in progranulin, a lysosomal and secreted protein with an unclear function. In a recent study, an unbiased metabolomics approach was used to discover that bis(monoacylglycero)phosphate (BMP) was deficient, and glucosylsphingosine accumulated, in tissues from GRN^−/− mice. These molecules were found to be defining molecular features of the condition, and could be used as target engagement biomarkers.²⁶

Another example of the use of discovery metabolomics is the identification of metabolites that are decreased in mice bearing PTEN-deficient tumors compared with nontumor-bearing controls, and their subsequent increase following dosing with the Class I phosphoinositide-3-kinase inhibitor pictilisib. These candidate metabolites were further evaluated in a Phase I dose-escalation clinical trial, in which time- and dose-dependent effects were observed in patients. This example highlights the feasibility and potential of discovery metabolomics as a strategy for identifying biomarkers that can be used to evaluate therapies.²⁷

A discovery metabolomics approach was crucial for discovering the oncometabolite 2-hydroxyglutarate (2HG) in isocitrate dehydrogenase mutant glioblastoma. Mutations in cytosolic isocitrate dehydrogenase 1 (IDH1) are common in a significant subset of primary human brain cancers. Using an unbiased metabolomics analysis, a significant elevation of a single metabolite peak was observed in cells with an R132H IDH1 mutation, which was identified as 2HG. Further data suggested that excess 2HG contributes to glioma formation and malignant progression and, thus, could be used as a biomarker for IDH mutations in gliomas, enabling more targeted treatment for specific individuals.^28-34 It is noteworthy that 2HG encompasses two enantiomers, namely D- and L-2HG, with D-2HG being the enantiomer associated with its oncometabolic attributes³⁵ Discriminating between chiral molecule enantiomers usually necessitates targeted strategies involving chiral chromatography.

Discovery metabolomics primarily provides relative or semiquantitative measurements of metabolites. This approach traditionally used untargeted methods and high-resolution, high-accuracy mass spectrometers that could measure thousands of metabolites in each sample. However, more recently, multiplexed assays performed with triple quadrupole mass spectrometers enable high-throughput, semiquantitative measurements of hundreds of selected metabolites per sample using only a few non-natural internal standards.^36-39 Regardless of the analytical approach, the results from unbiased analyses typically require further validation and confirmation using more quantitative metabolomics approaches.

Targeted metabolomics and biomarker validation

Targeted metabolomics analysis is a common technique used to monitor metabolites that could be potential disease biomarkers or therapeutic targets. This approach involves selecting a set of metabolites for analysis based on prior hypotheses, literature searches, or the results of unbiased discovery metabolomics. For example, after discovering a metabolomic signature associated with a lysosomal storage disorder in preclinical models, a selected set of metabolites was monitored as efficacy biomarkers for treatment in clinical studies using targeted metabolomics.^39-41 Similarly, in an animal model of methamphetamine addiction, specific alterations in lipid metabolites were discovered using unbiased discovery metabolomics, and these lipids and their metabolic pathways were then monitored to understand their mechanisms of action and monitor drug toxicity.⁴²

In clinical and epidemiological studies, which involve analyzing hundreds to thousands of samples, a targeted metabolomics approach is commonly used to ensure the results are reproducible over time and across different sample batches. Targeted measurements provide absolute quantification of metabolite concentration with high precision and accuracy, enabling comparisons between samples processed on different runs of the same or other instruments.⁴³

One way to gain a deeper understanding of the biology of a biomarker is to compare absolute values from multiple clinical studies. Targeted metabolomics can be used to conduct these comparisons, and a recent study utilized this approach to compare two independent studies. The results showed that concentrations of phosphatidylinositols were consistently higher in healthy participants than in patients with coronary artery disease. This discovery could help to develop a lipid biomarker panel that can be explored in future studies.⁴³

To conduct targeted analysis, triple quadrupole mass spectrometers are commonly used. Before sample preparation, stable, isotope-labeled internal standards for each metabolite of interest are usually added to the biological samples to normalize for sample preparation and detection variability. This technique generates more quantitative data compared with unbiased analysis. The nature of a targeted metabolomics approach allows for the development of high-throughput sample preparation and MS methods tailored to the analysis of selected metabolites.

Targeted analysis of biomarkers requires different degrees of analytical validation depending on the context of use of these measurements. For exploratory biomarkers intended for internal decision-making during drug development, a lower level of validation may be sufficient, such as when investigating the mechanism of action and selecting compounds, making go/no go decisions, or conducting proof-of-concept studies. However, when biomarkers are measured as primary and secondary endpoints in clinical trials, more extensive method validation and quality assurance are required, such as adhering to Good Clinical Practice/Good Clinical Laboratory Practice standards. Fit-for-purpose bioanalytical method validation can be conducted by following the US Food and Drug Administration’s (FDA) 2018 Bioanalytical Method Validation Guidance for Industry or the European Medicines Agency's 2011 Guideline on Bioanalytical Method Validation. If individual patient treatment or medical decisions are intended, biomarkers must be tested following Clinical Laboratory Improvement Amendments standards in the USA. This includes biomarkers used for clinical trial enrollment criteria, individual dose selection, and patient stratification. It is essential to review the appropriate development course with the relevant regulatory agency.

Fluxomics

Metabolomics provides information about the metabolites in a biological sample, but it does not capture the dynamic process of how those metabolites are processed. Fluxomics, also known as metabolic flux analysis, addresses this limitation by monitoring the flow of metabolites through metabolic pathways. Changes in the concentration of a metabolite in a disease state may indicate problems with degradation or increased biosynthesis. Fluxomics uses stable isotope-labeled metabolites as tracers to track their metabolic fate in biological systems. The concentration and labeling pattern of these metabolites vary based on the regulation of metabolic fluxes. By analyzing the masses and fragmentation patterns of the isotopically labeled metabolites, the position of the isotopes on their structure can be determined, and the regulation of biochemical reactions can be understood. Fluxes provide a direct metric of enzyme activity, which can aid in identifying enzymatic drug targets.^44-47

Metabolic flux analysis is a crucial tool for cancer metabolism research that can help identify unique metabolic characteristics of cancer cells and potential targets for new therapies. One example is the use of fluxomics to study cancer cell lines, which revealed that amplification of a metabolic enzyme called phosphoglycerate dehydrogenase (PHGDH) diverts glycolytic intermediates into the serine biosynthesis pathway, potentially contributing to tumor formation and development. In vitro studies showed that knocking down PHGDH can inhibit the growth of melanoma, breast cancer, and esophageal squamous cell carcinoma lines, indicating that PHGDH is a clinically actionable hallmark of cancer.^48,49

Fluxomics has a crucial role in immuno-oncology, where studies have demonstrated a close link between immune cell function and energy metabolism.⁵⁰ By targeting specific metabolic pathways of cancer cells, metabolic flux analysis has shown potential in modulating antitumor immunity and overcoming resistance to immune checkpoint blockade. One hallmark of cancer cell metabolism is the reliance on an exogenous supply of glutamine, a non-essential amino acid. Metabolic flux analysis revealed that a significant portion of glutamine-derived nitrogen and carbon is incorporated into the tricarboxylic acid (TCA) cycle to produce α-ketoglutarate, as well as to support the biosynthesis of amino acids, nucleotides, lipids, and polyamines essential for cell proliferation.^51,52 The inhibition of glutaminase, a mitochondrial enzyme that converts glutamine to glutamate, a precursor of α-ketoglutarate, is being evaluated in clinical trials for the treatment of various malignancies.^53-55

Although metabolic flux analyses are primarily conducted using cellular models, assessing metabolism in intact tumors is crucial to understand the combined effects of various intrinsic and extrinsic factors that cannot be fully captured in culture models. Recent studies introduced a new approach that uses stable isotope-labeled nutrients (such as [13C] glucose) to probe metabolic activity within intact tumors in vivo, in both mice and humans.⁵⁶ This approach offers the potential for novel insights to support targeted drug therapy that cannot be obtained by considering cellular models alone.

Finally, in addition to the utilization of stable isotopes, non-endogenous metabolites resembling endogenous compounds can be instrumental in tracing metabolite fate and understanding complex biological systems. As an example, researchers administered non-endogenous lipid 10Z-heptadecenoic acid (HDA, 17:1Δ10), akin to oleic acid, to rats to investigate its role in oleoylethanolamide biosynthesis. Using MS, they traced the fate of HDA, finding it mainly as a non-esterified fatty acid or glycerol ester in chylomicron components, such as triacylglycerol and phosphatidylcholine. Notably, HDA was metabolized into heptadecenoylethanolamide and its precursor N-10-heptadecenoyl-phosphatidylethanolamide. By contrast, CD36 transporter-null mice exhibited suppressed HDA absorption and heptadecenoylethanolamide formation. This suggests that exogenous HDA enters small-intestinal cells via a CD36-dependent pathway, leading to the production of fatty acyl ethanolamides.⁵⁷ This and analogous approaches underscore the effectiveness of using tracer metabolites, offering valuable insights into the rates of metabolite synthesis, disposition, and utihzation within a living organism^58-60 as well as probing enzymatic kinetics using cell-based models.^61,62

Spatial metabolomics

Spatial metabolomics allows for the visualization of the spatial composition of metabolites within different regions and subregions of biological tissue samples. This approach has many applications in drug discovery, including determining the spatial distribution and composition of endogenous metabolites in response to drug treatment, and investigating the absorption, distribution, and metabolism of drugs in animal tissue subregions.^39,63,64

Unlike traditional sample preparation methods, spatial metabolomics bypasses the homogenization step by using a focused beam or solvent droplets to irradiate a frozen tissue section, leading to the desorption of charged metabolites from the tissue. These metabolites are then guided into a mass spectrometer, and computational processing of the resulting data produces topographic maps reflecting the molecular composition of the tissue along its x and y axes.⁶⁶ Thus, spatial metabolomics provides a sort of molecular histology that can highlight metabolites in their natural environment without dissecting and processing the tissue.

Various ionization technologies can be used to obtain a metabolite image from a slice of frozen tissue, and coupling these with high-resolution and high-accuracy mass spectrometers allows for a discovery metabolomics approach. More recently, triple quadrupole instruments were used for targeted metabolomics, leading to highly sensitive and quantitative imaging measurements of metabolites and drug candidates in biological samples.⁶⁶

Spatial metabolomics has been used to determine the spatial pharmacodynamic response to drug treatments and to investigate drug absorption, distribution, and metabolism in specific organs or regions of animal tissues.⁶³ For example, a recent study examined the spatial localization of metabolite-based biomarkers of pathway engagement in response to enzyme replacement therapy (ERT) targeting the lysosomal enzyme iduronate 2-sulfatase (IDS).³⁹ Using imaging MS, the authors observed the enrichment of several species of gangliosides in the brains of IDS knockout (KO) mice, similar to results observed in the analysis of homogenized tissues. Treatment with an IDS ERT technology engineered to cross the blood–brain barrier could qualitatively reduce the accumulation of these ganglioside species throughout the brain regions. This specific localization of metabolic alterations occurring only in selected brain subregions was only possible using a spatial metabolomics approach (Figure 3).³⁹

Figure 3.

Schematic of a fluorescence-activated cell sorting (FACS) experiment used to isolate pure populations of neurons, astrocytes, microglia, and downstream endpoints analyzed to assess the cell type-specific distribution and efficacy of treatment in the brain. The levels of lipid biomarkers of lysosomal function, including gangliosides, were rescued in various cell populations isolated from brains of iduronate 2-sulfatase (IDS)-knockout (KO) mice treated with brain-penetrant IDS enzyme replacement therapy (ERT) compared with the animals treated with vehicle or commercial IDS ETR (idursulfase). Using spatial metabolomics, it was possible to visualize the response to eRT on the levels of gangliosides in select brain regions. Adapted from ³⁹.

An emerging research area is exploring the integration of spatial metabolomics and fluxomics. An illustrative instance involves the mapping of fractional fluxes across the brain, unveiling distinct biochemical pathways activated within tumor tissue and gauging the extent of metabolic heterogeneity.⁶⁷ This innovative methodology unearthed alterations spanning numerous anabolic pathways, a feat unattainable using traditional methods. Moreover, its potential application extends to diverse tissues and phenotypes in forthcoming investigations.

Post-analytical approaches

After completing the sample analyses, several post-analytical steps are necessary to obtain the final metabolomics data set. These typically include processing the raw data obtained from the instrument by removing technical variations and errors, normalizing the data, and filtering out irrelevant data. A table is then generated, reporting the abundance of identified or putative metabolites in each sample. Statistical analysis can then be performed to, for example, identify metabolites that show significant changes between experimental groups, such as upregulation or downregulation in response to treatment.

The next step is to identify the chemical structures of significant metabolites by comparing their mass spectra and retention times to known metabolite databases. These significant metabolites are then mapped onto metabolic pathways to understand the biological processes that are affected by the treatment.

For complex studies, biostatisticians and data scientists can assist in integrating the data set with study-specific variables, such as batch effects, and metadata, such as age, sex, fasting status, body-mass index, smoking status, and clinical records. This enables researchers to more accurately account for all possible sources of variation and potential noise to answer scientific questions of interest.

Metabolomics studies are often part of larger studies that include other ‘omics and molecular biology data, making functional metabolomics crucial for integrating data from different sources. Finally, targeted quantification approaches are used to validate significant metabolites by measuring their concentrations in a larger sample size, confirming the findings of the initial metabolomics experiment.

Metabolomics in drug discovery

Gene-based drug discovery

Recent advances in genetic screening technologies, such as genome-wide association studies (GWASs) and whole-genome sequencing studies, have facilitated the rapid identification of genetic variants associated with increased risk for many complex human diseases, including cancer, cardiovascular and respiratory diseases, and neurodegenerative disorders.^68,69 Although these genes and their related proteins could become targets for drug discovery, most of them have small effect sizes, and their functions remain largely unknown⁷⁰

Metabolomic studies provide a solution to this problem by using cellular and animal models to study the physiological and pathological functions of genes or variants of interest. These studies involve knocking down or knocking in a specific gene or variant in cellular and animal models. The resulting metabolite profile helps researchers to better understand the physiological and pathological functions of the gene or variant and can also serve as biomarkers for target and pathway engagement^39,71 (Figure 4a).

Figure 4.

Title. (a) Workflow of a typical translational metabolomics approach used to support gene-based drug discovery and the creation of connectivity maps. The maps can serve as a resource for finding connections between treatments sharing a mechanism of action (MoA) and genetic dysfunctions, thus matching potential treatments with high-value targets. (b) Retrotranslation workflow for metabolite-based drug discovery and phenotypic screening to first discover metabolomics signatures and then investigate drug MoA and efficacy aimed to rescue metabolomics phenotypes. Abbreviations: KI, knock-in; KO, knockout.

By analyzing metabolite profiles from cellular extracts or spent media with loss- or gain-of-function mutations in a particular gene using chemometric models, researchers can map gene–gene and gene–metabolome connectivity to ultimately understand gene function and the metabolite networks associated with it (Figure 4a). For example, deletion of two genes (TREM2 and PLCG2) in microglia cells resulted in very similar metabolite profiles, suggesting that the two genes encode proteins with some connectivity. The shared metabolite processing and accumulation defects between TREM2 and PLCG2 knockouts, combined with other signaling data, also suggested that PLCG2 activity is required for TREM2-dependent regulation of metabolism in human microglia.⁷²

Inborn errors of metabolism

Inborn errors of metabolism (IEM) are monogenic disorders that disrupt metabolic pathways and lead to specific metabolite deficiencies or accumulations. There are over 1000 well-characterized IEM conditions, collectively affecting ~1 in 1000 newborns.⁷³

Metabolomics has a crucial role in diagnosing and managing IEM by identifying specific metabolic perturbations associated with these diseases and aiding in the development of targeted therapies.⁷⁴ For example, metabolomics can be used to measure the levels of specific metabolites in blood or urine, which can help diagnose and monitor the progression of IEM.^75-77 While metabolomics approaches targeting specific metabolites are effective in screening for the most common IEMs, discovery metabolomics provides a more comprehensive coverage of metabolites and, therefore, is useful in identifying rare disorders that present with nonspecific clinical symptoms.^12,75-78

Metabolomics also provides insights into the molecular mechanisms underlying IEM and their associated pathologies. By identifying changes in the metabolite profile, metabolomics can help uncover the metabolic pathways that are affected by the disease and the downstream effects of these disruptions. This knowledge can then be used to develop targeted therapies that aim to restore normal metabolic function and alleviate symptoms associated with IEM.¹²

Lysosomal storage disorders and iduronate-2-sulfatase

Lysosomal storage diseases are a common form of IEM caused by lysosomal enzyme or protein defects. One example is mucopolysaccharidosis type II (MPS II), which is caused by a deficiency of IDS and characterized by the accumulation of glycosaminoglycans (GAGs). IDS KO mice are used as an experimental model, as they also accumulate GAGs in their tissues. Recently, a brain-penetrant IDS-ERT technology was developed that has shown promise in reducing GAGs in the brain. To assess treatment efficacy, researchers used fluorescence-activated cell sorting and metabolomics to isolate cell populations and analyze levels of metabolites. The results showed that the treatment not only reduced primary GAG accumulation but also improved lysosomal function by decreasing the levels of secondary lysosomal lipids, such as gangliosides, glucosylceramide, and BMP species. This demonstrates the potential of metabolomics in identifying changes in the metabolite profile that could be used to monitor treatment efficacy in lysosomal storage diseases.³⁹

Cancer and isocitrate dehydrogenase 1

Chronic and systemic loss of a particular metabolic activity resulting from IEMs can lead to the development of malignancy. Metabolic alterations observed in IEMs with an enhanced risk of cancer include mitochondrial dysfunction, toxin accumulation, oncometabolites generation, and metabolic diversion of flux.⁷⁹ Cancer cells exhibit significant metabolic differences from normal cells, including dysregulated glucose metabolism, altered fatty acid synthesis, and glutaminolysis, which can lead to therapeutic resistance.⁸⁰

Mutations in oncogenes and tumor suppressors accelerate and remodel selected metabolic pathways, providing the building blocks for rapid growth and control of signaling and epigenetic pathways that drive tumors. Therapeutic approaches targeting cellular metabolism have proven to be effective cancer treatments for decades, either alone or in combination with other therapies to enhance efficacy or overcome resistance.⁸¹

As discussed above, mutations in cytosolic IDH1 are a common feature of a major subset of primary human brain cancers. Cancer-associated IDH1 mutations result in the production of the oncometabolite 2HG, which can regulate epigenetic enzymes and contribute to the development of malignant brain tumors.^30,33,34 The unique biology of 2HG makes it a specific biomarker that can be used for diagnostic, prognostic, prediction, and pharmacodynamics assessment in clinical trials testing inhibitors against mutant IDH1 gliomas. Additionally, 2HG can be detected noninvasively by in-vivo magnetic resonance spectroscopy.^28,29,31,32

Alzheimer’s disease and TREM2

Large-scale GWASs have identified variants in Alzheimer's disease (AD) risk-associated genes involved in lipid metabolism, highlighting their therapeutic potential in neurodegeneration.^82,83 For example, loss-of-function variants in TREM2, which senses lipids and mediates myelin phagocytosis in microglia, increase the risk of AD.^84,85 A recent metabolomics study suggested that the loss of function of TREM2 impairs the ability of microglia to remove excess cholesterol from myelin, emphasizing the need to activate TREM2 for those carrying rare variants.⁷¹

Parkinson’s disease and LRRK2

Genetic risk factors for Parkinson's disease (PD), such as mutations in LRRK2,^86-89 have been identified through large-scale GWASs. Metabolomic profiling has shown that patients with the LRRK2-G2019S mutation have unique metabolomic profiles that distinguish them from patients with idiopathic PD. This differentiation could help predict which LRRK2 carriers will develop PD and support the development of clinical trials for potential therapies.⁹⁰ Environmental or lifestyle factors have also been linked to reduced or increased risk of PD,^91-93 and recent metabolomic analyses identified caffeine and its demethylation metabolites as prominent markers of resistance to PD.⁹⁴ Further investigation of the interplay between an individual's environmental exposures and their genome could lead to preventive and disease-modifying strategies for PD.⁹⁵

Metabolite-based drug discovery

Metabolomics not only supports gene-based drug discovery efforts, but can also serve as a starting point for drug discovery (Figure 4b). Metabolite-based drug discovery strategies include metabolome-wide association studies, phenotypic assays, and functional readouts of the human microbiome.

Metabolome-wide association studies

A growing number of metabolomics studies, including metabolome-wide association studies, measure metabolites in biospecimens from patients with a disease and then compare them with those from control individuals. The discovery of molecular signatures of the disease could lead to a better grasp of the molecular pathways accompanying or causing disease, helping generate novel hypotheses for target identification. To understand their functional relevance, the newly discovered metabolic alterations are then retro-translated and tested in more controlled experimental models, such as animals and cells, for target validation. In addition, besides providing mechanistic insight into the disease, identifying metabolites associated with disease development and progression could facilitate the discovery of predictive, diagnostic, and prognostic biomarkers, which could then be used for patient selection and to monitor therapeutic outcomes¹⁹

An example of a retro-translational metabolomics approach is discovering high serum levels of branched-chain amino acids (Ile, Leu, and Val) in patients with T2DM. Levels of these biomarkers can identify individuals at risk of developing T2DM as many as 15 years before disease onset and can be substantially more predictive compared with genetic data.^96-98 These amino acids specifically act on the mammalian rapamycin receptor and upregulate the same pathways and physiological processes as insulin.⁹⁸ High levels of amino acids can come from diet, yet they can also arise from the metabolism of the gut microbiome.⁹⁹

Phenotypic assays for drug screening and mechanism of action

A metabolomic-based phenotypic approach for drug discovery entails selecting the best therapeutic interventions either in an unbiased fashion or directed against a specific target of interest. It does so by monitoring the disease-relevant metabolomic phenotype as a biomarker of efficacy while simultaneously investigating the mechanism of action.^71,100-103

The metabolomic phenotyping approach for drug discovery could be summarized in three key steps: (i) identification of a specific metabolomic signature of the disease in humans or an animal model; (ii) generation of cellular models to recreate a metabolic phenotype similar to the disease condition. Lately, induced pluripotent stem cells containing a specific gene variant or subjected to treatments mirroring the disease phenotype have become prevalent;^71,104 and (iii) conducting a metabolomic phenotypic assay to identify different metabolic phenotypes corresponding to different mechanisms of action.

Mechanisms of actions that affect disease-relevant pathways could be further evaluated for in vivo proof-of-concept. Such cellular models could be additionally used to screen therapeutics for efficacy based on the alteration of metabolic phenotype. Ultimately, selected therapeutics will be tested in vivo for early development studies and eventually delivered to patients for clinical trials.

Metabolite-based functional readout of the human microbiome

The gut microbiome produces bioactive metabolites that affect metabolism, immune and inflammatory responses, and other physiological functions.¹⁰⁵ Microbiome-derived metabolites can signal to distant organs, enabling the gut microbiome to modulate host physiology, affecting key functions, such as metabolism, immune response, and inflammation^106-118 (Box 2). Metabolomics can help identify altered microbiome communities and functions in disease, providing potential drug targets. Modulating microbiome-derived metabolites by targeting gut bacteria genes or proteins, host enzymes and receptors, or the gut microbial population can lead to new target identification.^119,120 The gut microbiota also affects drug metabolism, bioavailability, efficacy, and toxicity, making functional metabolomic readouts of the microbiome valuable in early-stage clinical trials for precision medicine.¹²¹

Box 2. Microbiome-related metabolites.

Trimethylamine N-oxide

The gut-microbiome-derived metabolite trimethylamine N-oxide (TMAO) is one of the first examples of how metaorganismal pathways can affect disease susceptibility. The gut microbiome converts dietary precursors, such as carnitine and choline, into the intermediate trimethylamine, which is then delivered to the liver to be converted into TMAO.¹¹⁸ TMAO has been causally linked to the development of cardiovascular diseases, including atherosclerosis and thrombotic vascular disease,¹¹⁷ as well as AD.¹⁶⁵ Recent studies have shown that inhibiting this pathway through drug intervention can impede the development of arteriosclerosis in mice.

Short-chain fatty acids

Short-chain fatty acids (SCFAs) are primary metabolites produced by the bacterial fermentation of dietary fiber in the gastrointestinal tract. Microbiome-wide association studies of large population cohorts have highlighted associations between the gut microbiome, SCFAs, and complex traits, including T2DM and obesity.^166,167 SCFAs have also been implicated in microbiota–gut–brain communication. SCFAs could influence psychological functioning via interactions with G-protein-coupled receptors or histone deacetylases and exert their effects on the brain via direct humoral effects, indirect hormonal and immune pathways, and neural routes.¹⁶⁸

p-Cresol and indoles

p-Cresol and indoles are produced by gut bacteria during the breakdown of tyrosine/phenylalanine and tryptophan, respectively. After being absorbed, these metabolites undergo conjugation with sulfate. They can contribute to vascular and renal disease progression, suggesting new therapeutic targets for treating vascular calcification and stiffness in patients with chronic kidney disease.^169,170 Clinical trials are underway in oncology, dermatology, and gastroenterology to manipulate tryptophan metabolism via the indole pathway as a potential therapeutic strategy.¹⁷¹

An increase in p-cresol and its sulfated form, along with phenylacetylglutamine, has been observed in the serum of patients with PD. These are highly correlated with the severity of constipation and firmer stools in these patients.¹¹⁵ Intestinal bacteria primarily obtain energy through carbohydrate fermentation, generating butyrate and other SCFAs, or protein metabolism, generating amino acid metabolites, such as p-cresol and phenylacetylglutamine. The elevation of proteolytic bacterial metabolites and the decreased abundance of microbes with butyrate production capacity in PD serum suggest that PD is associated with a shift in colonic microbiota metabolism away from carbohydrate fermentation and toward proteolysis.

Elevated p-cresol has also been observed in the CSF of patients with PD and in autism spectrum disorder, another neurological condition with significant gut involvement.

Bile acids

Bile acids are the products of cholesterol metabolism, where primary bile acids are produced in the liver and further converted to secondary bile acids by gut bacteria. Besides aiding digestion and absorption, bile acids also have important signaling functions, which could affect the metabolism of various organs.

A recent study reported significantly lower levels of a primary bile acid, cholic acid, and higher levels of a secondary bile acid, deoxycholic acid, produced by gut bacteria, indicating the dysregulation of gut bacterial enzymatic activities. These changes were strongly associated with cognitive decline, a finding that was also replicated in serum and brain samples from an independent cohort. These results suggest the need for further research on gut dysbiosis and the gut–liver–brain axis in the pathogenesis of AD.¹⁷²

Biomarkers

Metabolite-based biomarkers

Metabolomics is a powerful tool in drug development, enabling us to understand how therapeutic interventions affect the metabolome. By measuring changes in endogenous metabolites, we can determine the mechanism of action of a drug, identify candidate drugs, and monitor drug response via pharmacodynamic biomarkers. Although metabolites might not always directly involve a pharmacological target, they can function as significant pathway engagement biomarkers, shedding light on the metabolic processes influenced by a drug target.^26,39,40 This capacity contributes to minimizing risk associated with drug targets and facilitates informed choices regarding leading candidates for subsequent development phases. Ultimately, metabolomics has the potential to be a vital component in shaping the design and success of clinical trials, providing deeper insights into drug metabolism and its implications for human health.

Translatable biomarkers

The lack of reliable biomarkers has been a major challenge to understanding the underlying pathophysiology and developing effective treatments for various diseases. In the case of neuropathic MPS II, current biomarkers, such as glycosaminoglycans (GAG) levels, in cerebrospinal fluid (CSF) and the periphery only provide limited information on disease progression and response to treatment. However, a metabolomics approach identified CSF accumulation of lysosomal lipids, reflecting the broad cellular dysfunction in the brain of patients with neuropathic MPS II.^39,40 Developing fluid-based biomarkers of primary storage products and secondary downstream pathology, such as lysosome dysfunction, could accelerate the field by providing a quantitative way to monitor patient responses to treatment, facilitating therapeutics development. Such biomarkers could be incorporated into the design of clinical trials to facilitate therapeutics development by supporting dose selection and duration of the treatment.^39,40

Pharmacometabolomic studies to further personalized drug treatment

Pharmacometabolomic studies analyze metabolic profiles with the aim of predicting how individual patients will respond to drug treatment.¹²² This approach is particularly useful for treating heterogeneous diseases, such as asthma, where metabolomic profiling can identify different subtypes of the disease and inform targeted therapeutic strategies.¹²³ For example, metabolomic analysis revealed the importance of altering the dosage of inhaled corticosteroids to prevent adrenal insufficiency and has shown how leukotriene inhibitors and bronchodilators affect patients' metabolomic profiles.^124,125 Pharmacometabolomics can also be used to monitor drug compliance, interactions, and adverse effects, as well as to assess environmental and lifestyle factors that might affect drug efficacy. Ultimately, analyzing metabolite profiles can help re-evaluate and interpret clinical trial data, increasing the success rate of future trials.

Metabolomics to support precision medicine

Population health studies have traditionally relied on human genomics to evaluate disease risk in large cohorts of participants. However, relying solely on genomics can sometimes leave gaps in our understanding of the development and progression of diseases. By incorporating metabolomics and other ‘omics, we can more accurately define an individual's health or wellness, predict disease progression, and tailor medical interventions, ultimately contributing to the advancement of precision medicine.⁴

Integrating metabolomics data with molecular biology and other ‘omics, including transcriptomics, genomics, metagenomics, and proteomics, allows for a more comprehensive understanding of the molecular pathways that are affected by a disease.^126,127 This deeper level of understanding can reveal alterations in metabolites and enable the stratification of patients into molecular subtypes based on their unique molecular heterogeneity. With time, these metabolomic data can aid patient selection and data interpretation, supporting informed decision-making in drug development.^89,128-152

The challenges of metabolomics research

Metabolomics research encounters various obstacles, which are prevalent in ‘omics discovery approaches, as well as analytical quantitative methodologies. Each metabolomics approach has its own challenges (Table 1), including:

Table 1.

Experimental MS-based metabolomics approaches for drug discovery and development

Approach	Advantages	Disadvantages
Discovery metabolomics	Enables biomarker discovery and target identification	Not fully quantitative: semiquantitative or qualitative
	Hypothesis generating, unbiased	Can miss low-abundance or unstable metabolites
	Comprehensive: no a priori selection of metabolites	Can require identification of unknowns
	Allows for data mining at any time to uncover unknown metabolites	High data volume: large amount of data can be generated
		More resource intensive: requires specialized equipment and expertise
		Complex data handling and analysis: requires advanced statistical and bioinformatics tools
Targeted metabolomics	Suitable for biomarker validation	Not comprehensive: limited to preselected known metabolites
	Mostly hypothesis driven	Unable to discover unknown metabolites that were not originally selected for analysis
	Quantitative measurements of specific metabolites of interest
		Can require development of specific analytical methods
	High sensitivity to detect low abundance metabolites	Dependence on availability of chemical standards for accurate quantification
	High-confidence chemical ID: easy to interpret
		Isotopically labeled standards used as internal standards can be expensive
	Scalable for high-throughput analysis
	Reduced data volume compared with discovery metabolomics
Fluxomics	Enable monitoring of selected metabolic reactions over time	Primarily used for cell or animal-based experiments
		Require high amount of tracer molecules, which can be expensive
	Useful for research and target identification
		Relatively complex and requires specialized expertise
Spatial metabolomics	Suitable for spatial localization of metabolites	Not fully quantitative
	High spatial resolution	Not high throughput
	Used to resolve tissue heterogeneity	High data volume: large amount of data can be generated
	Molecular histology
	Used to determine metabolites composition in subregions, which would otherwise be hard to dissect

• Standardization: there is a lack of standardization in sample collection, processing, and analysis, which can lead to variability and make it difficult to compare results across studies, resulting in inconsistencies and inaccuracies in data interpretation and hindering the development of reliable biomarkers.

• Metabolite identification: it can be challenging to identify and quantify metabolites, particularly for unknown or low-abundance compounds, which can affect the ability to make accurate biological interpretations. Reference databases can be limited, leading to uncertainties in metabolite identification and quantification.

• Data complexity: metabolomics approaches can generate large, complex data sets that can be difficult to interpret, especially for researchers with limited bioinformatics expertise. The analysis of large data sets requires sophisticated statistical tools and software, which can be time-consuming and require significant computational resources.

• Biological variability: there is inherent biological variability in metabolite levels resulting from factors such as age, sex, diet, and environmental exposures, which can complicate data interpretation and require careful study design and analysis. Additionally, biological samples are inherently heterogeneous, and metabolite concentrations can vary significantly depending on the tissue type and tissue region. This can make it challenging to identify consistent and reliable biomarkers across different sample types.

• Complexity of metabolomic networks: it is crucial to acknowledge that the application of metabolomics for target engagement might not consistently result in a direct and simple correlation. This complexity arises from the intricate interconnections within metabolic pathways. Nevertheless, even though metabolites might not always directly correspond to a pharmacological target, they can still function as noteworthy pathway engagement biomarkers, confirming that we are engaging the relevant biology. This capacity provides insights into the metabolic processes impacted by a drug target, which in turn contribute to reducing risks associated with drug targets and aiding the selection of promising candidates for subsequent development phases.

• Integration with other ‘omics data: integrating metabolomics data with other ‘omics data, such as genomics, transcriptomics, and proteomics, can provide a more comprehensive understanding of the metabolomic networks, biological processes, and disease mechanisms. However, data integration can be challenging because of differences in data types, and levels of biological complexity that make it difficult to establish a linear correlation between genes and metabolites (Figure 1).

Meeting these challenges demands sustained endeavors in standardization, collaboration, and innovative bioinformatics tool development. This is essential to enhance data quality and analysis and unlock the full potential of metabolomic-driven drug discovery. The Metabolomics Standards Initiative^153-155 and Lipidomics Standards Initiative^156-158 have introduced guidelines for reporting, covering experimental design, biological context, chemical analysis, and data processing across all phases of metabolomics and lipidomics analysis.

Emerging opportunities: publicly available data sets

Metabolomics provides a powerful approach to mining publicly available data sets, which can offer new opportunities for discovery and hypothesis generation. There are several popular repositories of metabolomics data sets, providing a valuable resource for researchers to access, including Metabolomics Workbench (www.metabolomicsworkbench.org) hosted by the National Institutes of Health, and MetaboLights (www.ebi.ac.uk/metabolights/) hosted by the European Bioinformatics Institute. Additionally, metabolomics data are now being incorporated into many large-scale databases collected by consortia, including the Alzheimer’s Disease Neuroimaging Initiative (https://adni.loni.usc.edu/), the Parkinson's Progression Markers Initiative (www.ppmi-info.org/), the UK Biobank (www.ukbiobank.ac.uk/), to name a few. Consortia provide a valuable resource for researchers to access and analyze metabolomics data sets generated by various research groups, allowing for the integration of multiple ‘omics data types to identify novel associations and gain deeper insights into the underlying biology of diseases.

Concluding remarks and future directions

Metabolomics is a potent technology that has broad applications in drug discovery and development. It enables the identification of changes in metabolic profiles, providing valuable biomarkers and shedding light on mechanisms of diseases and drug actions.

In the near future, the convergence of metabolomics with other technologies, such as gene-editing tools, cell-sorting and single-cell analysis, will reveal gene–metabolite networks, fundamental biology, and disease mechanisms, offering new venues for exploring drug targets and better understanding the mechanisms of action of drugs. The integration of metabolomics data with other ‘omics, including genomics, transcriptomics, proteomics, and microbiomics, will further enhance our understanding of health and disease and our ability to discover and develop novel treatments.

Although challenges in the use of these data remain, metabolomics-based biomarkers have the potential to have an increasingly vital role in disease risk stratification and precision medicine, allowing for the selection and customization of medical treatment based on individual metabolic profiles. This biomarker-driven approach to drug development can support the approval of new therapeutics and d-risk decision-making during the various phases of drug development.

Furthermore, metabolomics is likely to be increasingly used to provide a functional readout of genetic diversity, including common or rare mutations discovered using genomics approaches in large population studies. By integrating metabolites as longitudinal and phenotypic measurements with individual genome sequences, metabolomics promises to contribute to precision medicine, replacing current symptom-driven medicine and single disease-based prevention approaches.

In conclusion, based on what has been outlined in this review, we anticipate that metabolomics-driven drug discovery and development will become a key tool for researchers and scientists, supporting the identification of new disease targets and accelerating the approval of new treatments.

Highlights:

1. Metabolomics and lipidomics are important tools for measuring metabolites and lipids in biological systems, allowing for the identification of disease-related alterations and insights into disease etiology and pathological processes.

2. The application of metabolomics and lipidomics in drug discovery and development is increasing, as they provide functional readouts and composite measures of genetic susceptibility and environmental influences, leading to the identification of potential therapeutic targets.

3. Monitoring changes in metabolite and lipid levels or composition in relation to genetic variations can support human genetics and (meta)genomic approaches in drug discovery, enhancing the success rate of target identification and mechanism of action studies.

4. Metabolite- and lipid-based biomarkers are becoming essential in drug development, improving the understanding of drug efficacy and safety, and paving the way for precision medicine approaches.

5. This review covers the application of metabolomics and lipidomics throughout the drug discovery and development process, including basic research, target identification, mechanisms of action, and the development of clinically useful biomarkers.

6. The challenges of metabolomics research are discussed, along with emerging opportunities presented by publicly available datasets, offering new avenues for advancing the field.

Glossary

Chemometric models

mathematical, statistical, and other methods used to extract information from chemical systems by data-driven means

Discovery metabolomics

unbiased screening of as many metabolites as possible in a biological sample; normally provides a relative quantification of metabolite levels

Dysbiosis

‘imbalance’ in the gut microbial community, usually associated with disease

Endogenous metabolites

naturally produced substrates or products of the metabolic enzymes encoded in the human genome

Enzyme replacement therapy (ERT)

medical treatment that replaces an enzyme that is deficient or absent in the body; usually performed by giving the patient an intravenous (IV) infusion of a solution containing the enzyme

Exogenous metabolites

metabolite not naturally produced by an organism, such as a drug, chemical exposure, or food metabolite

Fluxomics

also called metabolic flux analysis; various approaches that seek to determine the rates of metabolic reactions within a biological entity

Fragmentation patterns

fragmentation is the dissociation of molecular ions formed from passing the molecules in the ionization chamber of a mass spectrometer. The fragments of a molecule cause a characteristic pattern in the mass spectrum, which could support the identification of their molecular structure

Glutaminolysis

process by which cells convert glutamine into TCA cycle metabolites

Idiopathic

disease or condition that arises spontaneously or for which the cause is unknown

Inborn errors of metabolism

diseases involving failure of the metabolic pathways involved in either the break down or storage of carbohydrates, fatty acids, and proteins

Monogenic disorders

traits or disorders caused by variation in a single gene

Oncometabolite

small-molecule components of normal metabolism the accumulation of which causes signaling dysregulation to establish a milieu that initiates carcinogenesis

Pharmacometabolomic studies

study of how differences in metabolites in an individual or subset of the population can be used to predict their varied responses to a drug or therapy

Spatial metabolomics

field of ‘omics research focused on the detection and interpretation of metabolites, lipids, drugs, and other small molecules in the spatial context of cells, tissues, organs, and organisms

Target engagement biomarkers

biomarkers proximal to the drug target that can be used to determine the dose required to fully engage the intended target and connect the drug target to physiological effects

Targeted metabolomic analysis

measurement of predefined metabolites using internal standards to obtain absolute quantification

Biographies

Giuseppe Astarita

With over 20 years of experience in academia and industry, Giuseppe Astarita is a distinguished expert in ‘omics technologies and biomarker discovery. Beyond research, Giuseppe's insights position him as a strategic advisor, emphasizing the pivotal role of ‘omics in drug discovery and development. A fervent advocate of technological innovation, his passion for uniting biology and technology drives his dedication to advancing scientific knowledge.

Rachel S. Kelly

Rachel Kelly is an assistant professor of medicine in the Channing Division of Network Medicine at Brigham and Women's Hospital and Harvard Medical School. Her primary focus is translational research, centered on metabolomic epidemiology and integrative ‘omics. She holds a specific interest in understanding the impact of early-life exposure on later-life disease risk mechanisms, particularly in common complex diseases.

Jessica Lasky-Su

Jessica Lasky-Su is a renowned leader in applying metabolomics to epidemiology, particularly in chronic diseases such as asthma. Her focus on ‘integrative metabolomics’ combines omics to study complex diseases and she is a driving force in the field. Initiatives such as STROBE-metabolomics and the Metabolomic Epidemiology Task Group showcase her influence. With a robust funding portfolio, Dr Lasky-Su is pivotal in multiomics, leading MGB-Biobank research.