Extensive characterization of xenobiotic metabolites remains a major challenge in mass spectrometry-based exposomics, as untargeted high-resolution mass spectrometry (HRMS) can detect extensive chemical diversity but often lacks sufficient structural annotation and prioritization strategies for unknown metabolites. Here, we developed a built-for-purpose HRMS-based workflow combining untargeted metabolomics, in silico metabolite prediction, structural annotation, and toxicity prioritization to characterize xenobiotic biotransformation products using Disperse Orange 61 (DO61) as a case study. Sprague-Dawley rats were exposed to DO61 via oral gavage, and liver, plasma, and urine samples were analyzed by ultra-high performance liquid chromatography (UHPLC) coupled to HRMS in positive and negative ionization modes. Statistically significant features were prioritized and screened against a BioTransformer-generated suspect library containing predicted Phase I, Phase II, reductive, and microbial transformations. HRMS-based annotation, supported by bromine isotope pattern analysis and MS/MS fragmentation, identified putative DO61 metabolites, while SIRIUS-derived molecular formula predictions provided additional confidence through fragmentation-based structural evaluation. The workflow experimentally confirmed reductive azo cleavage as an early DO61 transformation, producing two aromatic amine products, and revealed extensive downstream metabolism of the parent compound and both cleavage products. Computational toxicity prediction further prioritized confirmed and putatively annotated metabolites with potential mutagenic relevance for a future evaluation. This workflow demonstrates how integrating untargeted HRMS with computational approaches can facilitate metabolite discovery, improve structural annotation confidence, and prioritize metabolites for toxicological evaluation, providing a hypothesis discovery strategy for characterizing xenobiotic metabolism and supporting exposure and hazard assessment within the field of exposomics research.

Conventional analytical methods for detecting and quantifying phthalate esters (PAEs) in foods generally require lengthy sample pretreatment procedures and suffer from severe matrix interferences. To address these shortcomings, a fast analytical platform based on a thermal desorption unit that is directly interfaced with dielectric barrier discharge ionization mass spectrometry (TD-DBDI-MS) was developed. The integration of a heated nitrogen purge system effectively suppressed atmospheric background interference and improved signal stability. Moreover, a simplified liquid-liquid extraction pretreatment protocol was established that eliminates the need for centrifugation and solid-phase extraction (SPE) cleanup steps. The complete analysis cycle time per sample was reduced by approximately 55% compared to conventional GC-MS/LC-MS workflows, and each analysis by TD-DBDI-MS was completed in only 10 seconds. The method exhibited excellent linearity for four typical PAEs in the range of 0.5-200.0 μg/L with correlation coefficients (R2) > 0.998. The limits of detection (LOD) ranged from 0.11 to 0.52 μg/kg, spike recoveries were between 82% and 106%, and relative standard deviations (RSDs) were <10%. Application to 19 commercially available food samples revealed widespread PAE contamination, with butyl benzyl phthalate (BBP) being detected in every sample. Contamination levels followed the order premade dishes > milk > sugary beverages. PAE migration levels from plastic bag packaging were significantly higher than those from plastic bottle and carton packaging. Fortunately, the measured PAE concentrations were all well below the specific migration limits (SMLs) specified in China's GB 5009.271-2016 regulation and EU Regulation 10/2011. This rapid, efficient, and robust method provides a novel approach for large-scale screening of PAEs in complex food matrices.

A pulsed 193 nm deep ultraviolet (DUV) laser was used to ablate proteins and a noncovalently bound protein assembly deposited on surfaces. DUV laser ablation has been used for ablation of proteins from surfaces without fragmentation; however, DUV ablation of native proteins from surfaces without the loss of higher-order structure of proteins has not been demonstrated. Here, protein charge state distribution (CSD) and collision-induced unfolding (CIU) were used to probe the structure of DUV laser-ablated native proteins. Standard solutions containing native bovine serum albumin (BSA), jack-bean concanavalin A (ConA), and bovine carbonic anhydrase (CA) were deposited on a quartz microscope slide, laser-ablated, and captured for ion mobility-mass spectrometry (IM-MS) analysis. Observed CSD from native MS of conventionally extracted and DUV laser-ablated proteins were comparable. CIU plots for all charge states of laser-ablated proteins exhibited the same ion mobility features and comparable transition voltages (CIU50) as the extracted samples. These results suggest that DUV laser ablation can yield intact proteins with conserved higher-order structures.

The field of proteomics is one of the fastest-growing scientific disciplines in the world today. Beginning in 2020, ambitious market predictions anticipated proteomics to more than double in market size in 4-6 years and to even surpass the well-established genomic sciences industry in 8-12 years. While the market is still growing, it has not come close to matching the initial projections. To better understand the root cause of this, the author has collected field surveys and direct interviews of over 200 scientists actively involved in proteomics research today. While ultimately incomplete, it is worth noting that over 68% of respondents received their technical training in proteomics during either their doctoral or postdoctoral training. In stark contrast, only 4.3% of students obtained training during their undergraduate education. Given the 5.7-to-9.4-year average cycle for traditional doctoral and postdoctoral training in the life sciences, it appears the limiting factor in growing the proteomics field may be linked directly to current practices in training a qualified workforce.

The metabolic fitness of immune cells, reflected in part through glycosylation of surface proteins, is a key determinant of their activation, persistence, and effector responses. A recently developed high-throughput antibody capture cell array platform was further optimized to profile N-glycosylation across bulk cell and single-cell populations of peripheral blood mononuclear cells (PBMCs). Using antibodies against lymphocyte, monocyte, and dendritic cell markers (CD4+, CD8+, CD11c+, CD14,+ CD16+, CD3+, CD19+, CD56+, CD68+), N-glycosylation patterns of these immune cell subtypes were mapped across human PBMC cultures from healthy donors and those with pancreatic cancer. In this platform, PDMS stamps coated with targeted antibodies create patterned capture coordinates. Captured single cells are confirmed by brightfield imaging and selected via SoloCell software. Following PNGase F digestion and matrix application for MALDI-MSI analysis, N-glycans are acquired from a single laser burst per cell at a rate of 7.5 cells per second (18,000 cells in approximately 40 min). Sialic acid isomer stabilization by chemical amidation provided resolution of α2,6- versus α2,3-linked sialic acids and enabled detection of larger mono-, di-, tri- and tetra-sialylated glycans at the single-cell level. Bulk- and single-cell analyses showed concordant glycomic organization across immune cell subtypes. Bulk-level N-glycan profiles clearly distinguished lymphoid from myeloid lineages, and at single-cell resolution, healthy donor PBMCs similarly segregated into distinct subtype-specific glycomic clusters. In contrast, immune cell subtypes in pancreatic cancer PBMC converged into a common glycomic landscape, with marker-defined populations becoming highly intermixed. This convergence was accompanied by linkage-specific sialylation shifts, including a predominance of α2,6 sialylation and pronounced enrichment in CD3+ and CD8+ T cells. Together, these findings demonstrate that single-cell N-glycomic profiling extends bulk analysis to resolve subtype-specific differences in disease-associated immune N-glycome remodeling.

Matrix-assisted laser desorption ionization mass spectrometry imaging (MALDI-MSI) is the most widely used label-free method for mapping the molecular distribution of lipids and metabolites from tissue sections and cell cultures. To mitigate the problem of partially low MALDI ion yields for several of these analyte classes, different post ionization techniques have recently been developed─for example, MALDI-2, where a second UV-laser pulse intersects the initial particle plume to generate additional charges. Here, we show that the MALDI-2 signal intensities of important analyte classes can be further increased upon introducing reactive dopant vapor into the MALDI-2 ion source. Dopant assistance as an optional add-on was realized in prototype form on MALDI-2 ion sources of a Spectroglyph Injector-Orbitrap MSI setup and a timsTOF fleX MALDI-2 instrument. Acetone, anisole, toluene, 2-propanol, and chlorobenzene were tested as dopants. Acetone proved to be the most effective and robust dopant for the given setup and MALDI matrix, and thus was used for further parameter optimization on both instruments. We demonstrate that with dopant-assisted (DA-)MALDI-2, a signal increase of up to 5 orders of magnitude, as compared to regular MALDI, is, for instance, achieved for several classes of sterols, glycerophospholipids, and sphingolipids. Next to the increased intensity of signals for abundant lipids, the total amount of annotated signals (based on MS1) was also majorly increased by the use of dopants. We demonstrate the potential for biomedical imaging applications with the examples of murine spleen and cerebellum sections, analyzed in the positive- and negative-ion mode.

Mass spectrometry imaging (MSI) is now a routine modality for spatial molecular profiling and is increasingly combined with histology, multiplexed imaging, and spatial transcriptomics to build richer descriptions of tissue than any single technique can provide. The conceptual case for this multimodal paradigm was settled a decade ago, but the data infrastructure needed to operationalize it at scale has not kept up. Vendor binaries and the imzML exchange format keep MSI outside the OME-NGFF, Zarr, and SpatialData stacks on which the rest of spatial omics has converged, leaving each new experiment to rebuild ad hoc converters from scratch. Here, we present Thyra, an open-source Python library that converts raw vendor and imzML MSI data directly into the SpatialData format. A two-pass streaming sparse matrix engine writes the spectral matrix straight into the output Zarr store, keeping peak memory bounded near 200 MB regardless of data set size and supporting 100+ GB acquisitions. On a representative mouse brain MALDI data set, this yields three- to 4-fold storage reductions and an order-of-magnitude improvement in m/z range query latency, the bottleneck for ion image generation. Once converted, MSI data are directly composable with the scverse analysis ecosystem; we quantify this with an scverse readiness benchmark in which spatial autocorrelation, pixel clustering, and neighborhood enrichment, all standard operations in spatial transcriptomics, run as a one-liner on the converted object. A brief exploratory analysis of a skin biopsy from a psoriasis patient illustrates the analytical surface that the unified container exposes.

Time-of-flight m/z discrimination limits the effective analytical m/z range for Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry because optimal ion arrival time at the ICR cell varies with m/z1/2. Previously, an auxiliary RF waveform that allowed sequential m/z-dependent axial ejection from an external ion accumulation multipole resulted in mitigation of time-of-flight m/z discrimination, which improved ion transfer across a decade in the m/z range 200-2000. Here, a resonant RF axial ejection system based upon the same principle is implemented to further expand the m/z range that can be observed in a single mass spectrum. This new RF ejection system was installed without any change to the internal vacuum components. The RF ejection system has been significantly simplified over the initial implementation and installed and verified on several FT-ICR mass spectrometers in a multilab collaboration.

Top-down proteomics (TDP) enables comprehensive characterization of intact proteoforms, but the analysis of membrane proteoforms remains challenging because of their hydrophobicity and relatively low abundance. To address these challenges, we previously developed Azo, a photocleavable surfactant that enables efficient solubilization and top-down analysis of membrane proteins. Subsequently, we established an Azo-enabled TDP method permitting the comprehensive characterization of phospholamban (PLN), a transmembrane protein that plays crucial roles in calcium handling, directly from cardiac tissue. However, the original workflow required at least 10 mg of human cardiac tissue, making it unsuitable for biopsy-scale and other sample-limited studies. Herein, we developed a streamlined Azo-enabled one-pot extraction strategy for high-sensitivity top-down analysis of PLN that reduces the required tissue input to 1 mg while preserving analytical performance. Using online liquid chromatography-tandem mass spectrometry, we achieved highly reproducible detection, quantification, and post-translational modification localization of endogenous PLN proteoforms from 1 mg of human cardiac tissue. Notably, the streamlined strategy reduced the tissue required for PLN protein extraction 10-fold while retaining the same capacity for comprehensive characterization in the one-pot extraction. Collectively, these results establish a practical strategy for TDP of PLN proteoforms from biopsy-scale and other sample-limited myocardial specimens.

Coenzyme A is an essential cofactor synthesized from pantothenate, which carries activated acyl groups in cellular metabolic reactions. Coenzyme A participates in catabolism of carbohydrates, fats, and amino acids; biosynthesis of fatty acids, cholesterol, and heme; and protein modification, including acetylation and 4-phosphopantetheinylation. Despite CoA's critical functions, the regulation of CoA levels and the rate of CoA synthesis in different cell types and disease states are not well understood. Acyl-CoA species are challenging to measure due to instability, poor ionization, and the wide range of biochemical properties conferred by different acyl chain lengths; moreover, most current methods cannot measure CoA isotopic labeling, required to quantify the CoA synthesis rate or to measure concentration using isotope-labeled internal standards. Here, we describe a method to quantify the concentration and isotopic labeling of total CoA, defined as the sum of CoASH plus all acyl-CoA species. Acyl-CoA species are hydrolyzed using sodium hydroxide to remove acyl chains; then, CoA is derivatized with N-ethylmaleimide (NEM). Following protein precipitation and solid-phase extraction, samples are analyzed by liquid chromatography-mass spectrometry. This method is linear in a wide range that captures mouse tissue CoA levels, with accuracy within 15% error and precision below 15% relative standard deviation for standards and tissue samples. We applied this method to measure total CoA concentration in five mouse tissues and the total CoA synthesis rate in mouse liver via infusion of 13C-15N-pantothenate. Overall, this method offers a tractable approach to measure total CoA concentration and synthesis rate in health and disease.

Stable carbon isotope ratio measurements (δ13C) are traditionally performed by isotope ratio mass spectrometry (IRMS), but recent advances in high-resolution Orbitrap mass spectrometry (Orbitrap MS) enable molecular-level isotope ratio analyses without conversion to simple gases. Here, we systematically evaluate and compare three sample-introduction strategies for Orbitrap-MS-based δ13C determination of benzoic acid─dual-inlet (DI), single-inlet (SI), and automated HPLC flow-injection (column-free mode)─assessing how each governs measurement accuracy and precision. Analyses were performed using paired injections of an isotopically certified benzoic acid reference (IAEA-601) and a benzoic acid sample previously characterized by elemental analysis-IRMS (EA-IRMS), following a seven-block bracketing sequence (Ref-Smp-Ref-Smp-Ref-Smp-Ref) to minimize temporal drift and to generate within-run triplicates. Across all configurations, high analytical precision was achieved, within-block relative standard error (σRSE) consistently below approximately 0.6‰. Under DI conditions, replicate means closely approached the EA-IRMS benchmark (-27.77‰), with residuals of -0.40 to +0.50‰ on the Exploris 240 and -0.58 to -2.73‰ on the Q Exactive. SI analysis enabled the acquisition of a higher number of individual scans per block, contributing to lower σRSE (approximately 0.50-0.65‰) but exhibited larger accuracy offsets (residuals up to approximately 1-3‰), indicating greater susceptibility to systematic bias. Automated HPLC injection on the Exploris 240, optimized for concentration and AGC (50 μmol L-1 and 2.5 × 105, respectively), delivered δ13C values of 28.1‰, -27.19‰, and -27.79‰ (residuals of approximately 0.20-0.45‰) with no significant differences among replicate means (ANOVA/Tukey, 95% confidence level). Collectively, these findings confirm Orbitrap-MS as a reliable alternative to IRMS for δ13C analysis of benzoic acid, and reveal that the choice of injection strategy is the primary determinant of the accuracy-precision-throughput balance achievable with this platform.

Hydroxyl radical protein footprinting (HRPF) is a method for protein higher-order structure analysis that oxidizes solvent-accessible side chains of amino acids. One major limitation of HRPF is that results represent an ensemble average of all conformers in solution, preventing resolution of conformationally heterogeneous systems. Here, we established an inline hydrophobic interaction chromatography (HIC)-HRPF workflow that enables conformer-specific labeling while correcting for gradient-dependent differences in radical scavenging. Significant radical scavenging disparity exists between ammonium sulfate and phosphate buffers, and we compensated for this with 7 mM HEPES supplementation, restoring effective radical dose parity across the gradient. Using apo and holomyoglobin as a benchmark, we achieved baseline chromatographic separation of conformers. Multi-dose HRPF revealed conformer-specific oxidation patterns: apomyoglobin displayed enhanced labeling in peptide 81-97, specifically residue H83, consistent with destabilization of the F-helix and opening of the heme pocket, while holomyoglobin exhibited increased oxidation in peptide 120-154 corresponding to M132, E137, and L138, suggesting hinge-like rearrangements in the H-helix upon heme incorporation. These results validate inline HIC-HRPF as a robust, conformer-specific probe of higher-order structure and highlight the necessity of scavenging parity across the gradient for reliable structural interpretation. The method establishes a framework for extending HRPF to dynamic conformationally heterogeneous systems resolvable by HIC.

Recent advances in quantum machine learning (QML) offer new opportunities for chemical property prediction, yet their practical performance relative to established classical models remains poorly characterized on real-world datasets. Building on our previous benchmark of classical and deep learning models for collision cross-section (CCS) predictions, we present an empirical comparison of quantum kernel regression (QKR) and a variational quantum regressor (VQR) against classical machine learning (CML) techniques, including Random Forest (RF), support vector regression (SVR), Lasso, and a Voting Regressor (VR) ensemble. Using an aquatic dissolved organic matter (DOM) dataset with five molecular descriptors and five chemical classes, we systematically investigate QKR feature map repetitions and classical readout kernels, as well as VQR circuit depths and optimizer choices, to identify performance-optimal QML configurations. Regression performance is evaluated using the mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2) across training-set sizes up to 600 molecules per class. Overall, RF and Lasso deliver the most robust CCS predictions, achieving R2 ≥ 0.9 for lignins, carbohydrates, lipids, and proteins. Meanwhile, the optimized QKR architecture exhibits substantial gains as data increase and achieves competitive performance on chemically homogeneous subsets such as carbohydrates and lipids. In contrast, the VQR ansatz remains consistently weaker than both QKR and the classical baselines across all classes and data sizes. The results highlight the strong influence of dataset size, feature representation, and molecular heterogeneity on QML performance while also identifying conditions under which quantum methods may become competitive in future studies.