1.1. Overview of Experimental Design of Stable Isotope Tracing Experiments

Unlike untargeted metabolomics analysis, which proceeds without the requirement of a hypothesis or a focus on a specific metabolic pathway, stable isotope studies begin with the design of a tracing protocol, which is optimized largely based on the metabolic pathways of interest. This process demands the careful consideration of factors such as tracer selection, labeling protocol, sampling time points, and tissue-specific metabolic dynamics.

Formulation of research question: Building upon prior discoveries and working with a formulated hypothesis often simplifies the process. However, for discovery studies lacking a pre-established hypothesis, it is essential to conduct untargeted metabolomics or other types of omics (e.g., gene expression) studies before performing a stable isotope tracing analysis [7,8]. These initial investigations help identify dysregulated metabolic pathways and formulate specific research questions that the stable isotope tracing experiment aims to address. This step is crucial for guiding the design of an appropriate isotope tracing study based on the knowledge of biochemical reactions involved in the pathway(s) of interest. Factors such as cell culture conditions, animal models, or human subjects should also be considered based on the specific research questions to be addressed.

Selection of tracer(s): The choice of appropriate stable isotopes (e.g., ¹³C, ¹⁵N, ²H, ¹⁸O) and specific labeled compounds for tracing the desired metabolic pathway requires an understanding of the metabolic reactions. The choice of tracers is highly dependent on the metabolic enzymes of interest. Stable isotope-labeled tracers that closely mimic endogenous metabolites provide accurate insights into metabolic fluxes. Many types of cancer cells utilize glucose and glutamine [9,10], and hence, when there is not a well-formulated hypothesis or a specific metabolic enzyme to monitor, uniformly labeled ¹³C tracers (e.g., [U-¹³C₆]glucose or [U-¹³C₅]glutamine) are frequently used to study central carbon metabolism. These tracers allow researchers to monitor the utilization of glucose or glutamine carbons for the biosynthesis of downstream metabolites [2,3]. While uniformly labeled tracers provide a broad view of metabolic flux, singly labeled tracers may be a good option for monitoring particular metabolic enzyme activities. For example, position-specific ¹³C tracers, such as [1-¹³C₁]-pyruvate, result in unlabeled citrate via pyruvate dehydrogenase activity, while citrate labeled with a single ¹³C atom ([M+1] isotopologue) would indicate pyruvate carboxylase activity [11].

¹⁵N tracers are commonly used as stable isotope labels for tracking nitrogen metabolism, such as the incorporation of ¹⁵N into amino acids, nucleotides, or hexosamine. ²H tracers are primarily used for monitoring reactions involving isomerase reactions and dehydrogenase reactions with NADH or NADPH as co-factors [12,13]. ²H₂O tracing is also widely used for quantifying gluconeogenesis and pathways leading to the de novo biosynthesis of macromolecules (e.g., fatty acids, proteins, and DNA), both in vivo and in vitro [14]. To track metabolism involving the use of oxygen or the production of reactive oxygen species, the use of ¹⁸O₂ as the tracer can provide essential information [15,16]. Additionally, H₂[¹⁸O] as a stable isotope carrier of ¹⁸O was used not only for monitoring phosphoryl turnover but also for tracking oxygen exchange in other reactions, such as the tricarboxylic acid (TCA) cycle [17]. It is important to consider the compatibility of tracers with analytical techniques. For example, using an instrument capable of distinguishing between ¹³C and ¹⁵N isotopologues is beneficial when performing ¹³C and ¹⁵N dual isotopic labeling experiments. A comprehensive review detailing the choice of stable isotope tracers and the corresponding metabolic reactions to be monitored is available elsewhere [18].

Labeling protocol optimization: To optimize isotopic enrichment while minimizing non-specific effects (e.g., insulin fluctuations) and metabolic perturbations, it is essential to tailor conditions such as tracer concentration, the route and timing of tracer administration, and labeling duration according to the metabolic activities and kinetics of the pathways of interest and the specific experimental models to be examined (cells, tumor slices, or whole organisms) [2,3,19]. In animal studies, tracer administration procedures should be assessed as certain techniques, such as anesthesia, have been demonstrated to impact specific metabolic activities [20]. It is also important to optimize the tracer dose to achieve sufficient measurement sensitivity while minimizing metabolic perturbations that can lead to experimental artifacts [21]. For monitoring metabolic activities with a fast turnover rate, such as glycolysis, a bolus injection or short-term infusion of ¹³C tracers is sufficient to achieve high enrichments in measured metabolites, but for monitoring protein or lipid synthesis, alternative tracer administration methods (e.g., via drinking water or diet) are necessary due to the slower turnover rate of these pathways [22]. Pilot experiments may be performed to capture metabolic dynamics under various conditions, guiding the definition of sampling time points based on pathway kinetics. Sample collection strategies should be designed to minimize metabolic disruption and post-harvest metabolite turnover, which is often achieved through snap-freezing in liquid nitrogen [23].

Sample processing and analysis: While errors should be managed throughout the process, analytical errors can typically be evaluated and minimized by integrating internal standards or incorporating quality control measures, such as the analysis of unlabeled samples or labeled standards to detect unexpected measurement interferences. In contrast, pre-analytical errors, such as delays in sample freezing or improper storage can compromise metabolite integrity and present significant challenges for monitoring and control. Strict adherence to sample collection and handling protocols is essential to minimize pre-analytical errors, preserve stable isotopic labeling, and maintain metabolite integrity.

Metabolites or lipids are typically extracted using organic solvents such as methanol–water mixtures, acetonitrile–water mixtures, or isopropanol–water mixtures, as well as more hydrophobic solvents like chloroform or methyl tert-butyl ether (MTBE) for lipids. Analytical techniques such as mass spectrometry (MS) and nuclear magnetic resonance spectroscopy (NMR) are commonly employed to measure isotopologues or isotopomers, respectively. Computational tools are used for calculating isotopic enrichment, correcting for natural isotopic abundance, or performing metabolic flux analysis to identify differences between experimental groups.

Validation and integration: The conclusions drawn from stable isotope tracing experiments require validation using complementary methods like genetic or pharmacological interventions. These interventions allow for the direct manipulation of key enzyme activities within dysregulated metabolic pathways, thus confirming their significant roles in cancer biology, such as influencing cancer proliferation or metastasis. Additionally, integrating stable isotope tracing data with other omics datasets, such as gene expression profiles or metabolite pool size measurements, facilitates a thorough comprehension of metabolic regulation and network interactions.

1.2. Analysis of Isotope Labeling with Nuclear Magnetic Resonance (NMR)

NMR spectroscopy: NMR spectroscopy is a powerful tool for analyzing stable isotope labeling patterns [24]. In NMR analysis, metabolites are exposed to radiofrequency pulses, and nuclei within the metabolites absorb energy and transition between various energy states. This process generates frequencies that offer insights into the chemical environment of the nuclei, facilitating the identification and quantification of metabolites. NMR-based isotope tracing has roots dating back to as early as the 1970s [25,26,27]. Isotopically labeled metabolites exhibit distinct NMR signals compared to their unlabeled counterparts due to changes in nuclear spin. By comparing the NMR spectra of labeled and unlabeled samples, researchers can identify isotopically enriched metabolites and quantify the extent of labeling, which allows for monitoring the fate of stable isotope-labeled tracers. For instance, in glucose metabolism studies, the incorporation of ¹³C-labeled glucose into downstream metabolites, such as lactate, citrate, glutamate, or alanine, can be monitored using NMR. Furthermore, unlike MS, NMR enables researchers to identify the positions of labeled atoms within molecules by analyzing the spectral patterns and chemical shifts of NMR peaks, providing insights into specific metabolic pathways, which is crucial for understanding pathways like glucose metabolism in which the position of the labeled atom may be indicative of its origin. By measuring the incorporation of labeled precursors into metabolites of interest, researchers can calculate flux rates and assess the activity of specific metabolic pathways.

Magnetic resonance imaging: Various non-invasive and nondestructive imaging techniques have been developed based on the principles of NMR [28,29,30,31]. Traditional magnetic resonance imaging (MRI) essentially measures hydrogen atoms, which are abundant in biological organisms, especially in water and fat, and provides information about the shape and internal structure of soft tissues [32,33]. Due to sensitivity issues, in many instances, MRI cannot directly monitor other metabolites in tissues unless they are isolated and enriched. The spatial resolution of metabolism has been achieved by combining PET-CT scanning or MRI with metabolite extraction from dissected tumors, followed by NMR or MS-based analysis [34]. By introducing hyperpolarized ¹³C tracers, hyperpolarized MRI (HP-MRI) has improved sensitivity [35] and enables the non-radioactive and non-invasive monitoring of metabolic processes in vivo, offering real-time metabolic information with spatial resolution [28,29,30,31,36]. HP-MRI provides insights into regional metabolic heterogeneity and facilitates the characterization of metabolic phenotypes associated with diseases such as cancer. For instance, MRI has advanced the understanding of glutamine metabolism in tumor xenograft models, highlighting the heterogeneity of tumor metabolism [37,38]. It has been used to noninvasively monitor pyruvate metabolism in prostate and breast cancers in the clinical setting [39,40,41].

Advantages: NMR-based metabolomics is nondestructive, allowing for the repeated analysis of the same sample without significant sample loss [42]. In contrast, MS-based methods involve ionization and fragmentation, altering the sample irreversibly with each analysis. Additionally, since the intensity of an NMR signal correlates directly with the number of nuclei responsible for a particular resonance, NMR inherently yields quantitative data [43,44]. An underappreciated aspect of NMR is its ability to accurately quantify low-isotope enrichments (e.g., ~0.1%) that are below the noise threshold of typical MS measurements [45]. Furthermore, NMR’s instrument-independent chemical shifts contribute to greater reproducibility of metabolomics results across different laboratories, allowing for data exchange among instruments, especially those with comparable hardware, or even data from different laboratories and with quite variable skills [44,46].

Limitations: NMR-based metabolomics does have limitations. NMR typically exhibits lower sensitivity compared to MS-based methods, resulting in fewer identified metabolites [42,47,48]. At the moment, the availability of NMR-specific data processing software may be more limited compared to MS-based metabolomics [42,49,50]. Despite these drawbacks, NMR remains a valuable tool for metabolomic studies, particularly for its nondestructive nature, reproducibility, and capability for spatial metabolite imaging.

1.3. Analysis of Stable Isotope Labeling Using Mass Spectrometry (MS)

Mass spectrometry: In addition to NMR-based metabolomics, MS, especially high-resolution mass spectrometry (HRMS), is a powerful analytical technique widely employed for studying stable isotope labeling in metabolomics research [2,5,51]. MS measures the mass-to-charge ratio (m/z) of ions generated from molecules through ionization techniques such as electrospray ionization (ESI). Isotopically labeled metabolites exhibit distinct mass shifts compared to their unlabeled counterparts due to the incorporation of heavier isotopes, such as ¹³C, ¹⁵N, ²H, or ¹⁸O. Well-calibrated HRMS instruments with high mass accuracy and resolution ensure the reliable detection and quantification of isotopically enriched metabolites, even at low abundance levels. Most MS-based metabolomics approaches typically employ liquid chromatography mass spectrometry (LC-MS) coupled with electrospray ionization (ESI). However, alternative separation methods, such as gas chromatography (GC), have also been used [52,53,54], offering complementary capabilities for analyzing certain metabolites, such as volatile metabolites. Additionally, tandem mass spectrometry (MS/MS) is commonly employed in MS-based metabolomics to facilitate metabolite identification. It involves the fragmentation of precursor ions to provide structural information, which can offer insights into the position of stable isotopes within the metabolite molecules [55]. Comprehensive reviews of LC-MS metabolomics method development and recent advancements in the field have been previously documented [56,57,58].

Mass spectrometry imaging: Even though MS data typically lack spatial resolution, efforts to address this limitation have led to the development of mass spectrometry imaging (MSI) [59,60,61]. MSI typically utilizes matrix-assisted laser desorption ionization (MALDI) [60] or related techniques, such as desorption ionization electrospray (DESI) [61]. MSI has proven valuable in measuring metabolomic flux in both healthy and tumorous tissues, revealing spatial heterogeneity in metabolic phenotypes [62,63]. However, compared to MRI, MSI-based techniques typically require the isolation of tissue sections [62,63] and have not yet been widely applied for the direct measurement of in vivo flux, even though efforts are made to advance the in vivo applications [64,65,66].

Advantages: The major advantage of MS is that MS methods offer high sensitivity and resolution compared to NMR, enabling the detection and identification of thousands of metabolites from a complicated biological sample [42,67]. In addition, the MS can be coupled with a variety of separation platforms, such as LC, GC, ion chromatography (IC), supercritical fluid chromatography (SFC), and capillary electrophoresis (CE), which help to resolve individual chemical components and improve sensitivity.

Limitations: However, there are several weaknesses to MS approaches. In contrast to NMR, MS-based metabolomics typically tracks isotopologues, and MS1 scans cannot resolve the positional isotopic isomers of a given metabolite because MS detects the mass-to-charge ratio rather than the molecular structure. In contrast, NMR spectroscopy can provide information about the spatial arrangement of atoms within a molecule, allowing for the resolution of positional isotopic isomers. Additionally, MS itself is not inherently quantitative. The intensity of MS signals can be influenced by various experimental factors, such as ionization efficiency, matrix effects, and instrument settings, and hence, directly comparing datasets from different batches or across laboratories can be challenging [68,69,70]. However, it is worth noting that Clark et al. report qualitatively similar PCA plots in their interlaboratory study, indicating that while direct data comparison is often difficult, the overall trends observed in data across laboratories largely align [69]. Additionally, the impact of interlaboratory variables on the MS signal readout of isotopologues or isotopic isomers of the same metabolite is consistent, and hence, variations are minimized when labeling percentages, rather than relative intensities, are reported in the stable isotope-tracing analysis.

Because of these complementary capabilities of MS and NMR, efforts have been made to enable their coupling in metabolomics analysis with enhanced confidence in metabolite identification [49,71]. Publicly available software for modeling combined MS and NMR datasets has been developed, which provides improved precision when applied to estimate metabolic fluxes from stable isotope tracing experiments [72].

2.1. Exploring Cancer Metabolism through Glutamine Isotope Tracing

Stable isotope tracing utilizing ¹³C- or ¹⁵N-labeled glutamine, coupled with various analytical techniques, such as LC-MS, GC-MS, or NMR, has been widely used for quantitatively monitoring glutamine metabolism in cancer research (Figure 1). Glutamine plays a pivotal role in the survival and proliferation of cancer cells by serving as a crucial carbon and nitrogen source [9,115]. This multifaceted role contributes to various metabolic pathways essential for tumor growth. Glutamine is utilized in cancer cells to provide carbon for the synthesis of lipids and metabolites in the TCA cycle, as well as nitrogen for the biosynthesis of hexosamine, amino acids, and nucleotides, thus supporting the demands of cancer cell proliferation and growth [115,116,117]. Glutamine is converted to glutamate through glutaminolysis catalyzed by glutaminase (GLS). Subsequently, glutamate is further deaminated to α-ketoglutarate (αKG) through the action of enzymes such as glutamate dehydrogenase (GDH) or various transaminases, including glutamate-oxaloacetate transaminase (GOT), glutamate-pyruvate transaminase (GPT), and phosphoserine transaminase (PSAT) [9,115]. Once present in the TCA cycle, αKG can be oxidized to generate four-carbon units or undergo reductive carboxylation to isocitrate and citrate, eventually leading to the synthesis of fatty acids [115,117].

By using [U-¹³C₅]-glutamine, researchers can track the flux of carbons from glutaminolysis via GLS to oxidative and reductive TCA metabolites (Figure 1). Specifically, they can analyze the isotopic labeling patterns of metabolites, such as glutamate (M+5) and citrate (M+4) from oxidative TCA cycle flux and citrate (M+5) from reductive carboxylation flux, to determine the extent of glutamine-derived carbon incorporation [118]. [1-¹³C]glutamine can be applied to specifically confirm reductive carboxylation activity since glutamine C1 is lost as CO₂ in the oxidative metabolism of αKG. However, it is important to measure isotope incorporation in sink products such as palmitate or 2-hydroxyglutarate (2-HG) and apply comprehensive flux modeling to determine the direction and magnitude of the isocitrate dehydrogenase (IDH) flux since some labeling in citrate results from reversible exchange with αKG rather than net reductive flux [119]. Glutamine contains the following two nitrogens: 5-N or amide-N and 2-N or amine-N. Glutamine with ¹⁵N labeling at the five position ([5-¹⁵N]glutamine, or [Amide-¹⁵N]glutamine) can be used to monitor nitrogen incorporation into amino acids, such as asparagine, nucleotides, and hexosamine, while ([2-¹⁵N]glutamine, or [Amine-¹⁵N]glutamine leads to the ¹⁵N labeling at the two position of amino acids, such as alanine, aspartate, etc. 2-¹⁵N labeling in aspartate is subsequently incorporated into newly synthesized nucleotides [8,120]. Uniformly ¹⁵N-labeled glutamine ([U-¹⁵N₂]glutamine) offers a comprehensive ¹⁵N labeling strategy to simultaneously enrich a variety of downstream nitrogen-containing metabolites.

Metabolic flexibility: Stable isotope tracing with ¹³C- or ¹⁵N-labeled glutamine helps to understand the metabolic flexibility of cancer cells and provides insights into combination therapies that target multiple metabolic vulnerabilities simultaneously, thereby improving the effectiveness of antitumor treatments [121]. In one study, the use of a [U-¹³C₅]-glutamine tracing analysis of osteosarcoma cells enabled researchers to investigate the contribution of the reductive carboxylation of glutamine-derived α-ketoglutarate to aspartate production, which helped to explain why osteosarcoma cells with deficiencies in succinate dehydrogenase (SDH), also known as electron transport chain (ETC) complex 2, continue to produce and rely on aspartate synthesis for proliferation [122]. When [U-¹³C₅]-glutamine was used as the tracer, the knockout (KO) of SDH decreased the oxidative production of aspartate (M+4) but showed an increase in aspartate (M+3) derived from reductive carboxylation of glutamine-derived α-ketoglutarate in osteosarcoma cells [122]. The inhibition of ETC complex 1 increased reductive carboxylation and pyruvate carboxylation-derived aspartate production and rescued cell proliferation in osteosarcoma cells with SDS deficiency [122]. In another study, SDH-deficient cells have been shown to rely on glutamate–pyruvate transaminase (GPT2) for proliferation through GPT2-dependent α-ketoglutarate and alanine production [123], which was monitored through isotope tracing using [U-¹³C₅]glutamine and [2-¹⁵N]glutamine. GPT2 drives reductive TCA anaplerosis and regenerates NAD+ to support glycolysis [123]. [U-¹³C₅]glutamine tracing confirmed that the inhibition of GPT2 in SDH-deficient cells decreased the levels of citrate (M+5) and cytosolic malate (M+3), accompanied by halting cancer proliferation [123]. In another study, an unbiased screening approach was employed to elucidate the adaptive mechanisms of cancers in response to glutamine deprivation, and this approach facilitated the identification of ALDH18A1, which encodes P5CS, the rate-limiting enzyme in the proline biosynthetic pathway, as a gene that cells can downregulate in response to glutamine starvation [124]. The subsequent [U-¹³C₅]glutamine and [U-¹³C₆]glucose tracing analysis confirmed decreased proline synthesis, which enabled carbon to be redirected toward de novo glutamate and glutamine synthesis, supporting the proliferation of glutamine-restricted cancer cells [124].

Metabolic heterogeneity: It is acknowledged that tumors display significant levels of genomic and metabolic heterogeneity [125,126,127]. Stable isotope tracing is of great value in examining both inter- and intratumoral metabolic heterogeneity. For example, stable isotope tracing assays with [U-¹³C₅]glutamine in precision-cut slices of patient-derived xenografts (PDX), followed by GC-MS analysis, demonstrated the metabolic heterogeneity of human prostate tumors [128]. In another study, multi-isotope imaging mass spectrometry with isotope ratio mass spectrometry was used to analyze tumors isolated from mice bearing melanoma and malignant peripheral nerve sheath tumors injected with multiple tracers, including ¹⁵N-labeled glutamine [129]. The ¹⁵N labeling was monitored through an increase in the ¹²C¹⁵N⁻/¹²C¹⁴N⁻ ratio [129]. The results revealed tumor metabolic heterogeneity [129].

Metabolic requirements in tumor metastasis: CB-839 is a GLS inhibitor and reduces the conversion of glutamine to glutamate [130]. CB-839, when combined with metformin, results in both primary osteosarcoma growth inhibition and a notable reduction in metastatic outgrowth in vivo [131]. A [U-¹³C₆,¹⁵N₂]glutamine tracing analysis confirmed that the diminished M+4 or M+2 TCA intermediates without affecting M+5 or M+3 TCA metabolites, indicating a reduction in oxidative TCA cycle activity but no effect on reductive activity [131]. In another study, a gene expression analysis of breast cancer (BT474) growth in various tissues suggested the upregulation of genes related to lipid metabolism in breast tumors growing in the brain [7]. A subsequent [U-¹³C₆]glucose tracing analysis with GC-MS and MALDI-MSI confirmed the elevated fatty acid synthesis. The genetic and pharmacological inhibition of FASN suppresses breast cancer growth in the brain, indicating that fatty acid synthesis could potentially serve as a therapeutic target for breast cancer brain metastases [7].

Tumor microenvironment: Stable isotope tracing using [U-¹³C₅]glutamine in vitro and in vivo has been employed to investigate the interplay between tumor metabolism and immunogenicity within the tumor microenvironment (TME) and the association with immunogenicity [132]. The findings indicate distinct metabolic profiles between immunologically “hot” and “cold” melanoma tumors, with immunologically hot tumors exhibiting increased utilization of glutamine compared to immunologically cold tumors, both in vivo and in vitro [132]. The results not only highlight the potential significance of glutamine metabolism activity as a prognostic factor in melanoma but also offer insights for designing metabolic therapies to enhance immunotherapy efficacy [132]. In another study, a [U-¹³C₅]glutamine tracing analysis revealed that proline originates from glutamine metabolism in cancer-associated fibroblasts, supporting the production of pro-tumorigenic collagen [133]. Decreasing proline synthesis by reducing the level of pyrroline-5-carboxylate reductase 1 (PYCR1) in cancer-associated fibroblasts is sufficient to reduce tumor collagen production, tumor growth, and metastatic spread in vivo [133].

Drug target engagement: Advancements in understanding glutamine metabolism have paved the way for identifying potential drug targets aimed at disrupting cancer proliferation by targeting enzymes involved in glutaminolysis [9,115,134]. As of now, the primary drug targets for inhibiting the incorporation of glutamine carbons into the TCA cycle and nitrogens into nucleotide and amino acid synthesis are glutaminase (GLS) and amidotransferases. Researchers have monitored the efficacy of inhibiting various enzymes associated with the TCA cycle, glutaminolysis, amino acid synthesis, and nucleotide synthesis using isotopes of glutamine, such as [U-¹³C₅] and [5-¹⁵N]glutamine (Table 1) [98,118,131,135,136,137].

The effect of CB-839 on glutamine metabolism was confirmed by using [U-¹³C₅]glutamine or [U-¹³C₅,¹⁵N₂]glutamine in sarcoma cells [131,138], ovarian cancer cells [139], prostate cancer cells [140], glioblastoma cells [136], and patient-derived xenografts from renal cell carcinoma [118]. Additionally, HP-MRI using [5-¹³C,4,4-²H₂,5-¹⁵N]-L-glutamine was employed for monitoring glutamine metabolism in pancreatic cancer xenograft models and confirmed the reduced glutamine conversion to glutamate in vivo by CBP-839 [37].

6-Diazo-5-oxo-l-norleucine (DON), recognized as a glutamine antagonist, along with its prodrug, DRP-104, effectively inhibits the glutamine-dependent metabolism [141,142,143,144]. The impacts of DRP-104 on TCA cycle intermediates or nucleotide biosynthesis were verified through a [U-¹³C₅]glutamine or [U-¹³C₅,¹⁵N₂]glutamine flux analysis in mice bearing lymphoma EL4 tumors [143], a syngeneic model of pancreatic ductal adenocarcinoma (PDAC) [143], or Keap1 mutant lung cancer cells [145]. Specifically, the decrease in the enrichment of the M+4 TCA intermediates (e.g., succinate, fumarate, and malate) in DRP-104–treated tumors compared to the control indicates that DRP-104 impeded the influx of glutamine-derived glutamate into the TCA cycle [143].

Dihydroorotate dehydrogenase (DHODH), an enzyme involved in the de novo pyrimidine nucleotide synthesis pathway, has been specifically targeted in cancer cells with IDH mutations. An analysis of [5-¹⁵N]glutamine metabolism has revealed that the inhibition of DHODH effectively halts nucleotide synthesis, particularly of pyrimidine nucleotides, thereby impeding cancer cell proliferation [146].

2.2. The Use of Other Isotope Tracers for Investigating Cancer Metabolism

Glucose tracers: The Warburg effect is the hallmark of cancer and is characterized by the preferential utilization of glycolysis over oxidative phosphorylation, even in the presence of oxygen [151]. The significance of glucose in supporting the survival of proliferating cancer cells underscores the critical role of glucose isotope tracers, such as [U-¹³C₆]glucose, in elucidating dysregulated cancer metabolism and monitoring metabolic responses to anticancer treatment (Figure 2). By studying the incorporation of carbon from [U-¹³C₆]glucose into glycolysis, the TCA cycle, the PPP, amino acids such as serine, or newly synthesized fatty acids, researchers can gain valuable insights into metabolic dynamics (Figure 2 and Table 1). Similar to what has been discussed in Section 2.1, [U-¹³C₆]glucose tracing serves not only to elucidate tumor metabolism but also to explore the interplay with the surrounding tumor environment and cells. For example, cancer cells harboring mutant IDH are known to overproduce D-2-hydroxyglutarate (D-2HG), which has been shown, through the isotope-tracing of [U-¹³C₆]glucose and [U-¹³C₃]lactate, to modulate glycolysis in CD8+ T cells by inhibiting lactate dehydrogenase (LDH) [152]. In another study, [U-¹³C₆]glucose together with N-[1,2-¹³C₂]acetyl-d-glucosamine revealed that in a PDA tumor microenvironment with limited glutamine and glucose, the hexosamine salvage played an important role in supporting hexosamine biosynthesis and PDA tumor proliferation [153].

Open in a separate window

Figure 2

Schematic depicting the potential fates of ¹³C in [U-¹³C₆]glucose. Abbreviations: HK, hexose kinase; ACYL, ATP citrate lyase; IDH, isocitrate dehydrogenase; IDH mut, IDH mutant; UGDH, UDP-glucose dehydrogenase; GFAT, fructose-6-phosphate amidotransferase; PHGDH, phosphoglycerate dehydrogenase; LDH, lactate dehydrogenase; ALT, alanine transaminase; FASN, fatty acid synthase; GOT, glutamate-oxaloacetate transaminase 2; HMGCR, 3-hydroxy-3-methylglutaryl-CoA reductase; Ac, acetyl; G6P, glucose-6-phosphate; G1P, glucose-1-phosphate; F6P, fructose-6-phosphate; R5P, ribose-5-phosphate; PRPP, phosphoribosyl pyrophosphate; UDP-GlcA, UDP-glucuronate; UDP-GlcNAc, UDP-N-acetylglucosamine; DHAP, dihydroxyacetone phosphate; G3P, glycerol-3-phosphate; 3PG, 3-phosphoglycerate; Pyr, pyruvate; Lac, lactate; OAA, oxaloacetate; a-kG, alpha-ketoglutarate; 2-HG, 2-hydroxyglutarate.

Figure 1

Open in a separate window

Schematic depicting the potential fates of ¹³C and ¹⁵N in [U-¹³C₆, ¹⁵N₂]glutamine. Abbreviations: MDH, malate dehydrogenase; ACYL, ATP citrate lyase; IDH, isocitrate dehydrogenase; IDH mut, IDH mutant; GDH, glutamate dehydrogenase; GLS, glutaminase; GFAT, fructose-6-phosphate amidotransferase; P5CS, pyrroline-5-carboxylate synthase; PYCR, Δ1-pyrroline-5-carboxylate reductase; OAT, ornithine aminotransferase; ODC1, ornithine decarboxylase 1; GOT, glutamate-oxaloacetate transaminase 2; ASNS, asparagine synthetase; OAA, oxaloacetate; α-kG, alpha-ketoglutarate; 2-HG, 2-hydroxyglutarate; P5C, pyrroline-5-carboxylate.

M+n denotes the incorporation of n ¹³C- or ¹⁵N-heavy isotopes. For example, M+4 citrate indicates the citrate containing four ¹³C and two ¹²C. For metabolites containing both ¹³C and ¹⁵N, such as [¹³C₅, ¹⁵N₁]glutamate derived from [U-¹³C₅, ¹⁵N₂]glutamine, M+5 glutamate indicates glutamate with five ¹³C, while M+1 glutamate indicates glutamate with one ¹⁵N. This diagram summarizes the possible fates of both ¹³C and ¹⁵N, and for studies using glutamine tracer that only contains one isotope, such as ¹⁵N, but not ¹³C, the M+n would only indicate the incorporation of n ¹⁵N. The green line indicates oxidative TCA and related reactions, while the pink line indicates the reductive TCA and related reactions. Dashed lines indicate multi-step reactions.

M+n denotes the incorporation of n ¹³C. For simplicity, this schematic does not take into consideration complicated situations, such as tracer recycling or multiple rounds of the TCA cycle. Furthermore, this diagram can serve as a reference when downstream metabolites are utilized as tracers. For instance, when acetyl-CoA is labeled using [U-13C]glucose, a similar labeling pattern of acetylated proteins or lipids is expected as when acetyl-CoA is labeled using [¹³C₂]acetate. Dashed lines indicate multi-step reactions.

Fatty acid tracers: Due to the metabolic flexibility of many types of cancer, multiple tracers are often employed in the same study to gain a comprehensive understanding of metabolic adaptations. Certain cancer cell populations acquire the ability to utilize alternative nutrients as part of resistance mechanisms to chemotherapy. For example, resistance to cisplatin, a platinum-based chemotherapy drug, has been associated with increased fatty acid accumulation and beta-oxidation for energy production [154]. Fatty acid uptake and glucose anabolism were monitored using deuterium tracers ([D₃₁]-palmitic acid, [D₃₄]-oleic acid, or [D₇]-glucose) through hyperspectral stimulated Raman scattering imaging of C-D bonds in cisplatin-resistant and cisplatin-sensitive cells [154]. The results revealed a higher uptake of palmitic and oleic acid fatty acids and a decreased uptake of glucose in resistant cells. These findings underscore the reliance on fatty acid accumulation for energy production in resistant cells, thus presenting novel targets for drug development [154].

Nucleoside tracers: PDAC is highly lethal due to resistance to many chemotherapy treatments [155]. In one study, PDAC is characterized by elevated levels of uridine phosphorylase 1 (UPP1), an enzyme that cleaves uridine to ribose-1-phosphate and uracil. In vivo and in vitro isotope tracing using [¹³C₅]uridine with uniformly labeled ribose carbon demonstrated labeling of ATP (M+5), AMP (M+5), ADP (M+5), and NAD⁺ (M+5, M+10), confirming the utilization of uridine metabolism for the ribosylation of adenine. Additionally, ¹³C labeling of glycolytic, PPP, and TCA cycle intermediates was observed, indicating that uridine can serve as an alternative energy and carbon source [156].

2.3. Uncovering Mechanisms of Resistance to Cancer Treatments through Stable Isotope Tracing Analysis

Challenges in the development of cancer therapies: The development of cancer therapeutics is challenging, partly due to the high resistance of cancer cells to treatments and their metabolic plasticity [9,10,157,158,159,160,161,162]. To address this, researchers have turned to comparing metabolic profiles using isotope tracing between resistant and sensitive cancer cells. By employing isotope tracers, researchers can track the fate of labeled metabolites within cancer cells, revealing metabolic adaptations that contribute to treatment resistance. Another critical consideration in drug development is adverse effects either due to off-target activity or unintended actions in normal tissues [9,163,164]. As a result, there is a growing trend toward using multiple inhibitors targeting different metabolic enzymes at tolerable doses to achieve minimal adverse effects and enhance antitumor efficacy [165,166]. Insights gained from isotope tracing studies can inform the development of combination therapies that target both the genetic and metabolic vulnerabilities of cancer cells, thereby enhancing treatment efficacy and overcoming resistance. In this section, we explore studies employing stable isotope tracing to uncover resistance mechanisms in cancer cells, with a particular emphasis on the utilization of ¹³C- or ¹⁵N-labeled glutamine, since glucose tracing has been extensively studied.

Resistant mechanisms of GLS inhibitors: Despite being recognized as a potent and well-tolerated GLS inhibitor, CB-839 has shown mixed results in clinical trials [167,168]. While it has demonstrated efficacy in certain patients [167], a randomized clinical trial investigating its combination with cabozantinib in metastatic renal cell cancer (RCC) did not improve efficacy [168]. Similarly, during a phase II clinical trial involving patients with stage IV non-small cell lung cancer (NSCLC) tumors carrying loss-of-function mutations in KEAP1 (KEAPSAKE trial, NCT04265534), CB-839 did not show advantageous effects compared to standard-of-care immunotherapy. To understand the resistant mechanisms, [U-¹³C₅]glutamine tracing was performed in mice carrying liver tumors with the deletion of GLS1 and a reduced expression of GLS2 (Gls1KO/shGls2). The presence of a substantial fraction of the M+4 malate derived from [U-¹³C₅]glutamine suggests that other enzymes can utilize glutamine as an amide nitrogen donor and generate glutamate [121]. This finding led to the approach of synergistic inhibition of both glutaminases and compensatory amidotransferases to block glutamine catabolism and the proliferation of cancer cells in vitro and in vivo [121]. In another study, infusions of [5-¹⁵N]glutamine in clear cell RCC tumorgrafts unveiled sustained amidotransferase activity despite glutaminase inhibition by CB-839 [118]. When JHU-083, a prodrug of DON, was tested, [U-¹³C₅, ¹⁵N₂]glutamine tracing demonstrated a greater inhibition of the ¹⁵N labeling of metabolites catalyzed by amidotranferases than the ¹³C labeling of TCA cycle intermediates, suggesting that JHU-083 inhibited amidotransferases more effectively than GLS, subsequently resulting in more robust inhibition of tumor growth in both immunocompromised and immunocompetent mice than CB-839 [118].

Identifications of resistant cells to chemotherapy and radiotherapy: Glutamine tracing experiments were also performed to understand the resistant mechanisms of other types of cancer treatments. For example, [U-¹³C₅]glutamine or [¹⁵N₂]glutamine was injected intravenously into acute myeloid leukemia (AML)-bearing mice with vehicle or induction chemotherapy (iCT) for assessing labeled metabolites in AML cells or residual AML cells surviving iCT treatment [169]. The results revealed an increase in glutamine metabolism and ¹⁵N incorporation into pyrimidines in residual AML compared to the control AML cells [169]. The subsequent inhibition of de novo pyrimidine synthesis by brequinar (BRQ) post-iCT treatment significantly prolonged the survival of AML-bearing mice compared to iCT treatment alone, suggesting that glutamine metabolism drives AML chemoresistance by supporting pyrimidine synthesis [169]. Metabolomics and [U-¹³C₅]glutamine tracing analysis of ovarian cancer cells, which were sensitive or resistant to cisplatin treatment, revealed that glutamine metabolism is enhanced in cisplatin-resistant ovarian cancer cells [170]. The inhibition of glutaminase by compound 968 partially reversed the resistance to cisplatin treatment in these ovarian cancer cells [170]. In another study, metabolomics and a [U-¹³C₅]glutamine tracing analysis of prostate cancer cells sensitive or resistant to radiotherapy demonstrated that radioresistant prostate cancer cells and prostate cancer stem cells have a high glutamine demand, and reducing glutamine metabolism via glutamine depletion or the inhibition of critical regulators of glutamine utilization restored prostate cancer cell radiosensitization [171].

Mechanisms of resistance to immunotherapy: It is increasingly recognized that nutrient availability and metabolic activities in the tumor microenvironment play a role in determining immune cell function and antitumor immunity [145,172,173,174]. [U-¹³C₅]glutamine was used to monitor the glutamine uptake by dendritic cells, regulated by the transporter SLC38A2 [173]. Limited glutamine availability in the tumor microenvironment or impaired glutamine uptake due to SLC38A2 deletion compromised antitumor immunity [173]. In a separate study, inhibiting glutamine utilization by tumor cells enhanced antitumor T cell responses, and the combination of a GLS inhibitor, CB-839, with a glutamine antagonist, DRP-104, demonstrated a synergistic effect, enhancing the response to anti-PD1 checkpoint inhibitor therapy in Keap1 mutant lung tumors [145]. The [U-¹³C₅]glutamine tracing analysis suggests distinct actions of CB-839 and DRP-104 in vivo and specifically, that DRP-104 impairs tumor proliferation by inhibiting glutamine-dependent nucleotide synthesis rather than inhibiting glutaminolysis [145].

Identifications of resistant cells to hormone therapy: The contribution of metabolic alterations to hormonal therapy resistance in prostate cancer remains poorly understood. A [U-¹³C₅]glutamine tracing analysis played an important role in revealing the types of prostate cancer cells in which glutamine metabolism was more susceptible to CB-839 and pinpointing the isoform of GLS1 that was more sensitive to CB-839 [140]. These experiments also led to the discovery that GLS1 isoform switch is one of the key mechanisms in hormonal therapy resistance and disease progression [140].

Mechanisms of resistance to targeted therapy: Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs), which target early-stage lung cancer with EGFR mutations, have been among the most successful targeted cancer therapies [175,176,177]. However, the development of resistance to EGFR-TKIs poses a significant challenge in clinical management [178]. A [U-¹³C₅]glutamine tracing analysis of acquired EGFR-TKI-resistant lung cancer cell lines has unveiled a critical dependency on glutamine for supporting TCA cycle activity and glutathione (GSH) biosynthesis [179]. This finding offers valuable insights into the metabolic adaptations underlying EGFR-TKI resistance [179].