Hypoglycemia may be induced by a variety of physiologic and pathologic stimuli and can result in life-threatening consequences if untreated. However, hypoglycemia may also play a role in the purported health benefits of intermittent fasting and caloric restriction. Previously, we demonstrated that systemic administration of ricin toxin induced fatal hypoglycemia in mice. Here, we examine the metabolic landscape of the hypoglycemic state induced in the liver of mice by two different stimuli: systemic ricin administration and fasting. Each stimulus produced the same decrease in blood glucose and weight loss. The polar metabolome was studied using ¹H NMR, quantifying 59 specific metabolites, and untargeted LC-MS on approximately 5000 features. Results were analyzed by multivariate analyses, using both principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA), to identify global metabolic patterns, and by univariate analyses (ANOVA) to assess individual metabolites. The results demonstrated that while there were some similarities in the responses to the two stimuli including decreased glucose, ADP, and glutathione, they elicited distinct metabolic states. The metabolite showing the greatest difference was O-phosphocholine, elevated in ricin-treated animals and known to be affected by the pro-inflammatory cytokine TNF-α. Another difference was the alternative fuel source utilized, with fasting-induced hypoglycemia primarily ketotic, while the response to ricin-induced hypoglycemia involves protein and amino acid catabolism.

Capillary electrophoresis-mass spectrometry (CE-MS) offers unique analytical advantages for polar metabolite profiling but has remained underutilized in metabolomics relative to liquid chromatography-MS (LC-MS), in part due to challenges in managing migration time drift during data analysis. Here we introduce the use of indexed migration time (iMT) for easily managing this aspect of CE-MS data for metabolomics. Migration time indexing using a panel of stable isotope-labeled (SIL) amino acid reference standards, stored as an iRT database in Skyline, outperformed both uncorrected migration time and relative migration time (RMT) correction across three independent analytical batches spanning 90 samples from four biological matrices. The indexed migration time approach achieved sub-1% relative standard deviation (RSD) in migration index across batches, compared to up to ~15% RSD for uncorrected migration times. Additionally, we evaluate the use of single-point external calibration in Skyline for the purposes of metabolite quantification from complex matrices in order to ease the burden of translational metabolite quantification from metabolomics using high-resolution mass spectrometry (HRMS). Single-point external calibration using a biological matrix-based calibrator was benchmarked against a 13-point linear calibration curve across a panel of amino acids; above 1 μM, greater than 95% of back-calculated concentrations fell within ±20% of multi-point calibration. Application of the complete workflow to plasma, serum, urine, and NIST Standard Reference Material (SRM)-1950 demonstrated low inter-batch variability by principal components analysis, broad metabolite coverage across 126 quantifiable analytes, and strong quantitative concordance (Deming slope = 0.862, pseudo-R ² = 0.994, n = 64 analytes) with an independent comprehensive reference dataset for NIST SRM-1950. Together, these results establish a practical mCE-HRMS metabolomics workflow that bridges targeted and discovery metabolomics paradigms and lays the groundwork for single-point external calibration as a powerful tool for translational metabolomics.

BACKGROUND: The venous and lymphatic outflows from organs may contain molecular signatures related to tissue function, but deep biomolecular analyses that compare outflow composition are lacking. Here, using blood and lymph samples from a piglet model system and mass spectrometry analysis, we compare protein and lipid cargo from 9 venous and 3 lymph depots centered on intestinal outflow. METHODS: We obtained venous blood and plasma from piglets using new methods for dissection and sample collection. We applied mass spectrometry–based proteomics and lipidomics, using both a low-volume method and an additional proteomics sample preparation method, the Seer Proteograph XT high coverage method for the proteomics. RESULTS: We detected 622 proteins and 1315 lipids across lymph and plasma. With the additional Seer proteomics method, we detected a further 7771 proteins across a subset of samples. We observed both expected and novel enrichments of proteins, including CCL21 (C-C motif chemokine ligand 21) and IGFBP (insulin-like growth factor-binding protein) 7 as proteins strongly enriched in lymph. When comparing lymph depots, we found that thoracic duct lymph is distinct from lymph draining the proximal and distal small intestine, especially in their lipidomic profiles, possibly reflecting differences in dietary lipid uptake. By performing integrative multiomics of proteomics and lipidomics, we show that apoproteins, such as the related apoA1 (apoprotein A1) and apoA2 (apoprotein A2) proteins, correlate with different lipid profiles and may associate with distinct functions across the plasma depots. CONCLUSIONS: These data identify molecules and biomarkers selectively enriched in adjacent lymph and venous drainage depots from the gastrointestinal tract. The analyses and figures present in this work are expanded upon in an interactive companion Web application at gutveinlymphomics.com, facilitating access to our integrated multiomics and advancing understanding of biomolecular trends across the intestinal tract.