Pharmacological SERCA activation limits diet-induced steatohepatitis and restores liver metabolic function in mice.

Metabolic dysfunction-associated steatotic liver disease is the most common form of liver disease and poses significant health risks to patients who progress to metabolic dysfunction-associated steatohepatitis. Fatty acid overload alters endoplasmic reticulum (ER) calcium stores and induces mitochondrial oxidative stress in hepatocytes, leading to hepatocellular inflammation and apoptosis. Obese mice have impaired liver sarco/ER Ca2+-ATPase (SERCA) function, which normally maintains intracellular calcium homeostasis by transporting Ca2+ ions from the cytoplasm to the ER. We hypothesized that restoration of SERCA activity would improve diet-induced steatohepatitis in mice by limiting ER stress and mitochondrial dysfunction. WT and melanocortin-4 receptor KO (Mc4r-/-) mice were placed on either chow or Western diet (WD) for 8 weeks. Half of the WD-fed mice were administered CDN1163 to activate SERCA, which reduced liver fibrosis and inflammation. SERCA activation also restored glucose tolerance and insulin sensitivity, improved histological markers of metabolic dysfunction-associated steatohepatitis, increased expression of antioxidant enzymes, and decreased expression of oxidative stress and ER stress genes. CDN1163 decreased hepatic citric acid cycle flux and liver pyruvate cycling, enhanced expression of mitochondrial respiratory genes, and shifted hepatocellular [NADH]/[NAD+] and [NADPH]/[NADP+] ratios to a less oxidized state, which was associated with elevated PUFA content of liver lipids. In sum, the data demonstrate that pharmacological SERCA activation limits metabolic dysfunction-associated steatotic liver disease progression and prevents metabolic dysfunction induced by WD feeding in mice.

Program for Integration and Rapid Analysis of Mass Isotopomer Distributions (PIRAMID).

SUMMARY: The analysis of stable isotope labeling experiments requires accurate, efficient, and reproducible quantification of mass isotopomer distributions (MIDs), which is not a core feature of general-purpose metabolomics software tools that are optimized to quantify metabolite abundance. Here, we present PIRAMID (Program for Integration and Rapid Analysis of Mass Isotopomer Distributions), a MATLAB-based tool that addresses this need by offering a user-friendly, graphical user interface-driven program to automate the extraction of isotopic information from mass spectrometry (MS) datasets. This tool can simultaneously extract ion chromatograms for various metabolites from multiple data files in common vendor-agnostic file formats, locate chromatographic peaks based on a targeted list of characteristic ions and retention times, and integrate MIDs for each target ion. These MIDs can be corrected for natural isotopic background based on the user-defined molecular formula of each ion. PIRAMID offers support for datasets acquired from low- or high-resolution MS, and single (MS) or tandem (MS/MS) instruments. It also enables the analysis of single or dual labeling experiments using a variety of isotopes (i.e. 2H, 13C, 15N, 18O, 34S). DATA AVAILABILITY AND IMPLEMENTATION: MATLAB p-code files are freely available for non-commercial use and can be downloaded from https://mfa.vueinnovations.com/. Commercial licenses are also available. All the data presented in this publication are available under the "Help_menu" folder of the PIRAMID software.

Glucose-6-phosphatase catalytic subunit 2 negatively regulates glucose oxidation and insulin secretion in pancreatic ß-cells.

Elevated fasting blood glucose (FBG) is associated with increased risks of developing type 2 diabetes (T2D) and cardiovascular-associated mortality. G6PC2 is predominantly expressed in islets, encodes a glucose-6-phosphatase catalytic subunit that converts glucose-6-phosphate (G6P) to glucose, and has been linked with variations in FBG in genome-wide association studies. Deletion of G6pc2 in mice has been shown to lower FBG without affecting fasting plasma insulin levels in vivo. At 5 mM glucose, pancreatic islets from G6pc2 knockout (KO) mice exhibit no glucose cycling, increased glycolytic flux, and enhanced glucose-stimulated insulin secretion (GSIS). However, the broader effects of G6pc2 KO on ß-cell metabolism and redox regulation are unknown. Here we used CRISPR/Cas9 gene editing and metabolic flux analysis in ßTC3 cells, a murine pancreatic ß-cell line, to examine the role of G6pc2 in regulating glycolytic and mitochondrial fluxes. We found that deletion of G6pc2 led to ∼60% increases in glycolytic and citric acid cycle (CAC) fluxes at both 5 and 11 mM glucose concentrations. Furthermore, intracellular insulin content and GSIS were enhanced by approximately two-fold, along with increased cytosolic redox potential and reductive carboxylation flux. Normalization of fluxes relative to net glucose uptake revealed upregulation in two NADPH-producing pathways in the CAC. These results demonstrate that G6pc2 regulates GSIS by modulating not only glycolysis but also, independently, citric acid cycle activity in ß-cells. Overall, our findings implicate G6PC2 as a potential therapeutic target for enhancing insulin secretion and lowering FBG, which could benefit individuals with prediabetes, T2D, and obesity.

Oxidation of hyperpolarized [1-13 C]pyruvate in isolated rat kidneys.

Kidneys play a central role in numerous disorders but current imaging methods have limited utility to probe renal metabolism. Hyperpolarized (HP) 13 C magnetic resonance imaging is uniquely suited to provide metabolite-specific information about key biochemical pathways and it offers the further advantage that renal imaging is practical in humans. This study evaluated the feasibility of hyperpolarization examinations in a widely used model for analysis of renal physiology, the isolated kidney, which enables isolation of renal metabolism from the effects of other organs and validation of HP results by independent measurements. Isolated rat kidneys were supplied with either HP [1-13 C]pyruvate only or HP [1-13 C]pyruvate plus octanoate. Metabolic activity in both groups was confirmed by stable renal oxygen consumption. HP [1-13 C]pyruvate was readily metabolized to [13 C]bicarbonate, [1-13 C]lactate, and [1-13 C]alanine, detectable seconds after HP [1-13 C]pyruvate was injected. Octanoate suppressed but did not eliminate the production of HP [13 C]bicarbonate from [1-13 C]pyruvate. Steady-state flux analyses using non-HP 13 C substrates validated the utilization of HP [1-13 C]pyruvate, as observed by HP 13 C NMR. In the presence of octanoate, lactate is generated from a tricarboxylic acid cycle intermediate, oxaloacetate. The isolated rat kidney may serve as an excellent model for investigating and establishing new HP 13 C metabolic probes for future kidney imaging applications.

INCA 2.0: A tool for integrated, dynamic modeling of NMR- and MS-based isotopomer measurements and rigorous metabolic flux analysis.

Metabolic flux analysis (MFA) combines experimental measurements and computational modeling to determine biochemical reaction rates in live biological systems. Advancements in analytical instrumentation, such as nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS), have facilitated chemical separation and quantification of isotopically enriched metabolites. However, no software packages have been previously described that can integrate isotopomer measurements from both MS and NMR analytical platforms and have the flexibility to estimate metabolic fluxes from either isotopic steady-state or dynamic labeling experiments. By applying physiologically relevant cardiac and hepatic metabolic models to assess NMR isotopomer measurements, we herein test and validate new modeling capabilities of our enhanced flux analysis software tool, INCA 2.0. We demonstrate that INCA 2.0 can simulate and regress steady-state 13C NMR datasets from perfused hearts with an accuracy comparable to other established flux assessment tools. Furthermore, by simulating the infusion of three different 13C acetate tracers, we show that MFA based on dynamic 13C NMR measurements can more precisely resolve cardiac fluxes compared to isotopically steady-state flux analysis. Finally, we show that estimation of hepatic fluxes using combined 13C NMR and MS datasets improves the precision of estimated fluxes by up to 50%. Overall, our results illustrate how the recently added NMR data modeling capabilities of INCA 2.0 can enable entirely new experimental designs that lead to improved flux resolution and can be applied to a wide range of biological systems and measurement time courses.

Toxicant-Induced Metabolic Alterations in Lipid and Amino Acid Pathways Are Predictive of Acute Liver Toxicity in Rats.

Liver disease and disorders associated with aberrant hepatocyte metabolism can be initiated via drug and environmental toxicant exposures. In this study, we tested the hypothesis that gene and metabolic profiling can reveal commonalities in liver response to different toxicants and provide the capability to identify early signatures of acute liver toxicity. We used Sprague Dawley rats and three classical hepatotoxicants: acetaminophen (2 g/kg), bromobenzene (0.4 g/kg), and carbon tetrachloride (0.3 g/kg), to identify early perturbations in liver metabolism after a single acute exposure dose. We measured changes in liver genes and plasma metabolites at two time points (5 and 10 h) and used genome-scale metabolic models to identify commonalities in liver responses across the three toxicants. We found strong correlations for gene and metabolic profiles between the toxicants, indicative of similarities in the liver response to toxicity. We identified several injury-specific pathways in lipid and amino acid metabolism that changed similarly across the three toxicants. Our findings suggest that several plasma metabolites in lipid and amino acid metabolism are strongly associated with the progression of liver toxicity, and as such, could be targeted and clinically assessed for their potential as early predictors of acute liver toxicity.

Biochemical and metabolic characterization of a G6PC2 inhibitor.

Three glucose-6-phosphatase catalytic subunits, that hydrolyze glucose-6-phosphate (G6P) to glucose and inorganic phosphate, have been identified, designated G6PC1-3, but only G6PC1 and G6PC2 have been implicated in the regulation of fasting blood glucose (FBG). Elevated FBG has been associated with multiple adverse clinical outcomes, including increased risk for type 2 diabetes and various cancers. Therefore, G6PC1 and G6PC2 inhibitors that lower FBG may be of prophylactic value for the prevention of multiple conditions. The studies described here characterize a G6PC2 inhibitor, designated VU0945627, previously identified as Compound 3. We show that VU0945627 preferentially inhibits human G6PC2 versus human G6PC1 but activates human G6PC3. VU0945627 is a mixed G6PC2 inhibitor, increasing the Km but reducing the Vmax for G6P hydrolysis. PyRx virtual docking to an AlphaFold2-derived G6PC2 structural model suggests VU0945627 binds two sites in human G6PC2. Mutation of residues in these sites reduces the inhibitory effect of VU0945627. VU0945627 does not inhibit mouse G6PC2 despite its 84% sequence identity with human G6PC2. Mutagenesis studies suggest this lack of inhibition of mouse G6PC2 is due, in part, to a change in residue 318 from histidine in human G6PC2 to proline in mouse G6PC2. Surprisingly, VU0945627 still inhibited glucose cycling in the mouse islet-derived ßTC-3 cell line. Studies using intact mouse liver microsomes and PyRx docking suggest that this observation can be explained by an ability of VU0945627 to also inhibit the G6P transporter SLC37A4. These data will inform future computational modeling studies designed to identify G6PC isoform-specific inhibitors.

Bacteria- and fungus-derived PAMPs induce innate immune memory via similar functional, metabolic, and transcriptional adaptations.

Exposure to pathogen-associated molecular patterns (PAMPs) induces an augmented, broad-spectrum antimicrobial response to subsequent infection, a phenomenon termed innate immune memory. This study examined the effects of treatment with ß-glucan, a fungus-derived dectin-1 ligand, or monophosphoryl lipid A (MPLA), a bacteria-derived Toll-like receptor 4 ligand, on innate immune memory with a focus on identifying common cellular and molecular pathways activated by these diverse PAMPs. Treatment with either PAMP prepared the innate immune system to respond more robustly to Pseudomonas aeruginosa infection in vivo by facilitating mobilization of innate leukocytes into blood, recruitment of leukocytes to the site of infection, augmentation of microbial clearance, and attenuation of cytokine production. Examination of macrophages ex vivo showed amplification of metabolism, phagocytosis, and respiratory burst after treatment with either agent, although MPLA more robustly augmented these activities and more effectively facilitated killing of bacteria. Both agents activated gene expression pathways in macrophages that control inflammation, antimicrobial functions, and protein synthesis and suppressed pathways regulating cell division. ß-glucan treatment minimally altered macrophage differential gene expression in response to lipopolysaccharide (LPS) challenge, whereas MPLA attenuated the magnitude of the LPS-induced transcriptional response, especially cytokine gene expression. These results show that ß-glucan and MPLA similarly augment the innate response to infection in vivo. Yet, MPLA more potently induces alterations in macrophage metabolism, antimicrobial functions, gene transcription and the response to LPS.

Immunoresponsive Gene 1 Facilitates TLR4-agonist-Induced Augmentation of Innate Antimicrobial Immunity.

Treatment with the toll-like receptor (TLR) 4 agonist monophosphoryl lipid A (MPLA) conditions innate immunocytes to respond robustly to subsequent infection, a phenotype termed innate immune memory. Our published studies show that metabolic reprogramming of macrophages is a prominent feature of the memory phenotype. We undertook studies to define the functional contributions of tricarboxylic acid (TCA) cycle reprogramming to innate immune memory. We observed that priming of wild type (WT) mice with MPLA potently facilitated accumulation of the TCA cycle metabolite itaconate at sites of infection and enhanced microbial clearance. Augmentation of itaconate accumulation and microbial clearance was ablated in immuneresponsive gene 1 (Irg1) -deficient mice. We further observed that MPLA potently induces expression of Irg1 and accumulation of itaconate in macrophages. Compared to WT macrophages, the ability of Irg1-deficient macrophages to kill Pseudomonas aeruginosa was impaired. We further observed that itaconate is directly antimicrobial against P. aeruginosa at pH 5, which is characteristic of the phagolysosome, and is facilitated by reactive oxygen species. MPLA-induced augmentation of glycolysis, oxidative phosphorylation and accumulation of the TCA cycle metabolites succinate and malate was decreased in Irg1 KO macrophages compared to WT controls. RNA sequencing revealed suppressed transcription of genes associated with phagolysosome function and increased expression of genes associated with cytokine production and chemotaxis in Irg1 deficient macrophages. This study identifies a contribution of itaconate to MPLA-induced augmentation of innate antimicrobial immunity via facilitation of microbial killing as well as impact on metabolic and transcriptional adaptations.

In vivo2H/13C flux analysis in metabolism research.

Identifying the factors and mechanisms that regulate metabolism under normal and diseased states requires methods to quantify metabolic fluxes of live tissues within their physiological milieu. A number of recent developments have expanded the reach and depth of isotope-based in vivo flux analysis, which have in turn challenged existing dogmas in metabolism research. First, minimally invasive techniques of intravenous isotope infusion and sampling have advanced in vivo metabolic tracer studies in animal models and human subjects. Second, recent breakthroughs in analytical instrumentation have expanded the scope of isotope labeling measurements and reduced sample volume requirements. Third, innovative modeling approaches and publicly available software tools have facilitated rigorous analysis of sophisticated experimental designs involving multiple tracers and expansive metabolomics datasets. These developments have enabled comprehensive in vivo quantification of metabolic fluxes in specific tissues and have set the stage for integrated multi-tissue flux assays.

Multitissue 2H/13C flux analysis reveals reciprocal upregulation of renal gluconeogenesis in hepatic PEPCK-C-knockout mice.

The liver is the major source of glucose production during fasting under normal physiological conditions. However, the kidney may also contribute to maintaining glucose homeostasis in certain circumstances. To test the ability of the kidney to compensate for impaired hepatic glucose production in vivo, we developed a stable isotope approach to simultaneously quantify gluconeogenic and oxidative metabolic fluxes in the liver and kidney. Hepatic gluconeogenesis from phosphoenolpyruvate was disrupted via liver-specific knockout of cytosolic phosphoenolpyruvate carboxykinase (PEPCK-C; KO). 2H/13C isotopes were infused in fasted KO and WT littermate mice, and fluxes were estimated from isotopic measurements of tissue and plasma metabolites using a multicompartment metabolic model. Hepatic gluconeogenesis and glucose production were reduced in KO mice, yet whole-body glucose production and arterial glucose were unaffected. Glucose homeostasis was maintained by a compensatory rise in renal glucose production and gluconeogenesis. Renal oxidative metabolic fluxes of KO mice increased to sustain the energetic and metabolic demands of elevated gluconeogenesis. These results show the reciprocity of the liver and kidney in maintaining glucose homeostasis by coordinated regulation of gluconeogenic flux through PEPCK-C. Combining stable isotopes with mathematical modeling provides a versatile platform to assess multitissue metabolism in various genetic, pathophysiological, physiological, and pharmacological settings.

In Vivo Estimates of Liver Metabolic Flux Assessed by 13C-Propionate and 13C-Lactate Are Impacted by Tracer Recycling and Equilibrium Assumptions.

Isotope-based assessment of metabolic flux is achieved through a judicious balance of measurements and assumptions. Recent publications debate the validity of key assumptions used to model stable isotope labeling of liver metabolism in vivo. Here, we examine the controversy surrounding estimates of liver citric acid cycle and gluconeogenesis fluxes using a flexible modeling platform that enables rigorous testing of standard assumptions. Fasted C57BL/6J mice are infused with [13C3]lactate or [13C3]propionate isotopes, and hepatic fluxes are regressed using models with gradually increasing complexity and relaxed assumptions. We confirm that liver pyruvate cycling fluxes are incongruent between different 13C tracers in models with conventional assumptions. When models are expanded to include more labeling measurements and fewer constraining assumptions, however, liver pyruvate cycling is significant, and inconsistencies in hepatic flux estimates using [13C3]lactate and [13C3]propionate isotopes emanate, in part, from peripheral tracer recycling and incomplete isotope equilibration within the citric acid cycle.

CRISPR-Mediated Isogenic Cell-SELEX Approach for Generating Highly Specific Aptamers Against Native Membrane Proteins.

INTRODUCTION: The generation of affinity reagents that bind native membrane proteins with high specificity remains challenging. Most in vitro selection paradigms utilize different cell types for positive and negative rounds of selection (where the positive selection is against a cell that expresses the desired membrane protein and the negative selection is against a cell that lacks the protein). However, this strategy can yield affinity reagents that bind unintended membrane proteins on the target cells. To address this issue, we developed a systematic evolution of ligands by exponential enrichment (SELEX) scheme that utilizes isogenic pairs of cells generated via CRISPR techniques. METHODS: Using a Caco-2 epithelial cell line with constitutive Cas9 expression, we knocked out the SLC2A1 gene (encoding the GLUT1 glucose transporter) via lipofection with synthetic gRNAs. Cell-SELEX rounds were carried out against wild-type and GLUT1-null cells using a single-strand DNA (ssDNA) library. Next-generation sequencing (NGS) was used to quantify enrichment of prospective binders to the wild-type cells. RESULTS: 10 rounds of cell-SELEX were conducted via simultaneous exposure of ssDNA pools to wild-type and GLUT1-null Caco-2 cells under continuous perfusion. The top binders identified from NGS were validated by flow cytometry and immunostaining for their specificity to the GLUT1 receptor. CONCLUSIONS: Our data indicate that highly specific aptamers can be isolated with a SELEX strategy that utilizes isogenic cell lines. This approach may be broadly useful for generating affinity reagents that selectively bind to membrane proteins in their native conformations on the cell surface.

G6PC2 confers protection against hypoglycemia upon ketogenic diet feeding and prolonged fasting.

OBJECTIVE: G6PC2 is predominantly expressed in pancreatic islet beta cells. G6PC2 hydrolyzes glucose-6-phosphate to glucose and inorganic phosphate, thereby creating a futile substrate cycle that opposes the action of glucokinase. This substrate cycle determines the sensitivity of glucose-stimulated insulin secretion to glucose and hence regulates fasting blood glucose (FBG) but not fasting plasma insulin (FPI) levels. Our objective was to explore the physiological benefit this cycle confers. METHODS: We investigated the response of wild type (WT) and G6pc2 knockout (KO) mice to changes in nutrition. RESULTS: Pancreatic G6pc2 expression was little changed by ketogenic diet feeding but was inhibited by 24 hr fasting and strongly induced by high fat feeding. When challenged with either a ketogenic diet or 24 hr fasting, blood glucose fell to 70 mg/dl or less in G6pc2 KO but not WT mice, suggesting that G6PC2 may have evolved, in part, to prevent hypoglycemia. Prolonged ketogenic diet feeding reduced the effect of G6pc2 deletion on FBG. The hyperglycemia associated with high fat feeding was partially blunted in G6pc2 KO mice, suggesting that under these conditions the presence of G6PC2 is detrimental. As expected, FPI changed but did not differ between WT and KO mice in response to fasting, ketogenic and high fat feeding. CONCLUSIONS: Since elevated FBG levels are associated with increased risk for cardiovascular-associated mortality (CAM), these studies suggest that, while G6PC2 inhibitors would be useful for lowering FBG and the risk of CAM, partial inhibition will be important to avoid the risk of hypoglycemia.

Mechanism-based identification of plasma metabolites associated with liver toxicity.

Early diagnosis of liver injuries caused by drugs or occupational exposures is necessary to enable effective treatments and prevent liver failure. Whereas histopathology remains the gold standard for assessing hepatotoxicity in animals, plasma aminotransferase levels are the primary measures for monitoring liver dysfunction in humans. In this study, using Sprague Dawley rats, we investigated whether integrated analyses of transcriptomic and metabolomic data with genome-scale metabolic models (GSMs) could identify early indicators of injury and provide new insights into the mechanisms of hepatotoxicity. We obtained concurrent measurements of gene-expression changes in the liver and kidneys, and expression changes along with metabolic profiles in the plasma and urine, from rats 5 or 10 h after exposing them to one of two classical hepatotoxicants, acetaminophen (2 g/kg) or bromobenzene (0.4 g/kg). Global multivariate analyses revealed that gene-expression changes in the liver and metabolic profiles in the plasma and urine of toxicant-treated animals differed from those of controls, even at time points much earlier than changes detected by conventional markers of liver injury. Furthermore, clustering analysis revealed that both the gene-expression changes in the liver and the metabolic profiles in the plasma induced by the two hepatotoxicants were highly correlated, indicating commonalities in the liver toxicity response. Systematic GSM-based analyses yielded metabolites associated with the mechanisms of toxicity and identified several lipid and amino acid metabolism pathways that were activated by both toxicants and those uniquely activated by each. Our findings suggest that several metabolite alterations, which are strongly associated with the mechanisms of toxicity and occur within injury-specific pathways (e.g., of bile acid and fatty acid metabolism), could be targeted and clinically assessed for their potential as early indicators of liver damage.

Pancreatic islet beta cell-specific deletion of G6pc2 reduces fasting blood glucose.

The G6PC1, G6PC2 and G6PC3 genes encode distinct glucose-6-phosphatase catalytic subunit (G6PC) isoforms. In mice, germline deletion of G6pc2 lowers fasting blood glucose (FBG) without affecting fasting plasma insulin (FPI) while, in isolated islets, glucose-6-phosphatase activity and glucose cycling are abolished and glucose-stimulated insulin secretion (GSIS) is enhanced at submaximal but not high glucose. These observations are all consistent with a model in which G6PC2 regulates the sensitivity of GSIS to glucose by opposing the action of glucokinase. G6PC2 is highly expressed in human and mouse islet beta cells however, various studies have shown trace G6PC2 expression in multiple tissues raising the possibility that G6PC2 also affects FBG through non-islet cell actions. Using real-time PCR we show here that expression of G6pc1 and/or G6pc3 are much greater than G6pc2 in peripheral tissues, whereas G6pc2 expression is much higher than G6pc3 in both pancreas and islets with G6pc1 expression not detected. In adult mice, beta cell-specific deletion of G6pc2 was sufficient to reduce FBG without changing FPI. In addition, electronic health record-derived phenotype analyses showed no association between G6PC2 expression and phenotypes clearly unrelated to islet function in humans. Finally, we show that germline G6pc2 deletion enhances glycolysis in mouse islets and that glucose cycling can also be detected in human islets. These observations are all consistent with a mechanism by which G6PC2 action in islets is sufficient to regulate the sensitivity of GSIS to glucose and hence influence FBG without affecting FPI.

Network Modeling of Liver Metabolism to Predict Plasma Metabolite Changes During Short-Term Fasting in the Laboratory Rat.

The liver-a central metabolic organ that integrates whole-body metabolism to maintain glucose and fatty-acid regulation, and detoxify ammonia-is susceptible to injuries induced by drugs and toxic substances. Although plasma metabolite profiles are increasingly investigated for their potential to detect liver injury earlier than current clinical markers, their utility may be compromised because such profiles are affected by the nutritional state and the physiological state of the animal, and by contributions from extrahepatic sources. To tease apart the contributions of liver and non-liver sources to alterations in plasma metabolite profiles, here we sought to computationally isolate the plasma metabolite changes originating in the liver during short-term fasting. We used a constraint-based metabolic modeling approach to integrate central carbon fluxes measured in our study, and physiological flux boundary conditions gathered from the literature, into a genome-scale model of rat liver metabolism. We then measured plasma metabolite profiles in rats fasted for 5-7 or 10-13 h to test our model predictions. Our computational model accounted for two-thirds of the observed directions of change (an increase or decrease) in plasma metabolites, indicating their origin in the liver. Specifically, our work suggests that changes in plasma lipid metabolites, which are reliably predicted by our liver metabolism model, are key features of short-term fasting. Our approach provides a mechanistic model for identifying plasma metabolite changes originating in the liver.