Journal of Proteome Research最新文献

Histone deacetylase-containingcomplexes play centralroles inchromatin regulation, yet the structural organization of their intrinsicallydisordered regions (IDRs) remains poorly understood. Here, we identifythe previously uncharacterized protein C16orf87, which we renamedMHAP1, as a binding partner of HDAC2 and MIER1. Because all threeproteins contain extensive IDRs that remain difficult to study usingsingle structural analysis methods, we employed an integrative structuralapproach combining cross-linking mass spectrometry (XL-MS) with integrativestructural modeling to define the architecture of the complex. Theresulting integrative structural model reveals that the previouslyuncharacterized C-terminal IDR of HDAC2 acts as a key interactionhub, engaging both the ELM2 domain of MIER1 and the N- and C-terminalregions of MHAP1. These findings provide the first structural frameworkfor the interactions among HDAC2:MIER1:MHAP1 and demonstrate how intrinsicallydisordered regions can mediate assembly of a chromatin-associateddeacetylase complex. Finally, this research further demonstrates theability of XL-MS to facilitate integrative structural modeling ofprotein complexes. Mass spectrometry data may be accessed throughProteomeXchange via PXD082280 and the MassIVE repository MSV000102752.

Although aberrant glycosylation is closely associatedwith lungcancer (LC), salivary glycopatterns and their discriminatory valueacross healthy individuals, benign pulmonary disease, and LC remaininsufficiently characterized. We profiled saliva from 307 participantsusing a lectin microarray, comprising a development cohort of 227participants and an independent test cohort of 80 participants. Fivetask-specific LASSO-logistic regression models were constructed todistinguish pulmonary disease, LC, and LC subtypes. A separate Nom-LCmodel combined consensus lectin features selected by LASSO and supportvector machine recursive feature elimination with clinical variables.Salivary glycopatterns differed among healthy volunteers, benign pulmonarydisease, and LC groups, and the five task-specific models showed discriminatoryperformance across pulmonary disease, LC, and exploratory subtypeclassification tasks. The resulting Nom-LC model incorporated smokinghistory and four lectin signals (HHL, PNA, RCA120, and PWM), yieldingAUCs of 0.908 and 0.907 in the training and internal validation sets.In the independent test cohort, the AUC was 0.903 (95% CI, 0.831–0.976);at the prespecified cutoff, sensitivity, specificity, and accuracywere 86.36%, 94.44%, and 90.00%, respectively. These findings supportfurther evaluation of salivary glycopatterns as an auxiliary toolfor LC risk stratification, but prospective multicenter validationis required.

Theaim of the present study was to identify serum proteins associatedwith atherosclerosis and its comorbidities. The 3-year follow-up included216 patients subjected to coronary angiography. Gensini score wasused to quantify coronary atherosclerosis, and peripheral atherosclerosiswas assessed by B-mode ultrasound. Serum proteins were profiled usingan Explorer antibody microarray (Full Moon Biosystems, USA). In-houseELISA or commercial ELISA kits were used to validate the microarraydata in the total cohort for the top five proteins from the atheroscleroticprofile. Only ICAD (DNA fragmentation factor subunit, DFF45/ICAD)was linked to type 2 diabetes and the presence of coronary arterystenosis regardless of the severity of stenosis. CD53 was also associatedwith coronary stenosis. IL-6 was associated with diabetes, and GSTmu(glutathione S-transferase Mu3) was associated with brachiocephalicstenosis ≥ 60% and with femoral and coronary arterial lesions.Only GRM1 (glutamate metabotropic receptor 1) showed inverse associationswith coronary stenosis and cardiovascular outcomes (r = −0.24and P = 0.025 and r = −0.31 and P = 0.007, respectively). Moreover,GRM1 was inversely associated with femoral lesions ≥ 30%. Thus,we identified previously unreported associations of the ICAD, CD53,GRM1, and GSTmu proteins with atherosclerosis, diabetes, and nonfatalcardiovascular outcomes.

The estimation of the biological sex of archeologicalremains iscrucial information in bioarcheology and forensic anthropology. Inrecent years, proteomics based on molecular sexual dimorphism hasemerged as a preferred method, particularly because of its minimallyinvasive approach for extracting amelogenin X and Y proteins fromtooth enamel. However, there is an increasing demand to acceleratethis process and facilitate the analysis of large archeological assemblages.This study presents a novel high-throughput targeted paleoproteomicsmethod for biological sex estimation using MALDI-CASI-FTICR mass spectrometry.This approach combines the strengths of existing methods, includingultrahigh resolution, significantly reduced processing times, targetedanalysis, and scalability to large archeological sample sets. Themethod was initially validated on modern individuals with known sexand subsequently applied to 130 adult and juvenile individuals frommedieval Great Moravia (present-day Czech Republic). Biological sexwas successfully estimated for all but one of the individuals. Theresults not only provide a more efficient biological sex estimationbut also help to resolve a few errors in sex assessment previouslyencountered with osteomorphological and tooth morphometric techniques.The implementation of this method significantly improves the accuracyand efficiency of biological sex estimation, offering a powerful toolfor anthropological research.

Ontogenetic shifts in diet are well documented in snakesand areincreasingly linked to age-related venom variation. The common adder, Vipera berus, exhibits a dietary transition from predominantlyectothermic prey in its early life to increasingly incorporating endothermicprey as an adult. Here, we investigate whether this dietary shiftis reflected in age-related changes in the venom composition and bioactivityof V. berus. Venoms from captive-bred V. berus from Germany were obtained and pooled across five agegroups, from neonates to adults. Venom profiles were assessed by SDS-PAGEand genome-guided shotgun proteomics, with quantification based onnormalized spectral abundance factors (NSAF) using a toxin-gene cataloguegenerated from a novel V. berus genome assembly.In parallel, we assayed general protease and PLA₂ activities,as well as FXa-, thrombin-, and plasmin-like activities, and cytotoxicitytoward mammalian cell lines. We identified two distinct age-relatedvenom phenotypes (ontotypes): an svMP/CTL-rich ontotype A (≤1year) and an svSP/PLA₂-rich ontotype B (≥2 years).Functionally, protease activity decreased with age, whereas thrombin-like,plasmin-like and PLA₂ activities, and cytotoxicity, increased.Our findings indicate an ontogenetic shift in composition and activitiesof V. berus venom that parallels dietary transitionsand potentially reflect adaptation to differing prey physiologies.

Metabolic reprogramming fuels cancer progression, butwhether commonmetabolic patterns exist across diverse malignancies remains incompletelyunderstood. To address this, we integrated large-scale proteomic,transcriptomic [The Cancer Genome Atlas (TCGA)], and spatial transcriptomic(Spatial Meta-Transcriptome Database) data sets comprising 3226 samplesacross 24 human cancer types. Utilizing a highly controlled, predominantlypatient-matched design to minimize background noise, we characterizedthe pan-cancer metabolic landscape via reaction-level functional taskscoring, pathway enrichment, multiomics integration, network modeling,survival analysis, and microenvironmental coupling. Consistently upregulatedglycan biosynthesis pathways─particularly fucosylation andsialylation─and broadly enhanced nucleotide metabolism highlightedconserved programs potentially associated with tumor growth, adaptation,and immune evasion. These signatures proved robust against variationsin tumor purity across the TCGA cohorts. Furthermore, gene coexpressionnetwork modeling prioritized specific hub genes associated with thispan-cancer restructuring. Exploratory high-resolution mapping alignedtumor immunogenicity with specific glycan pathways. Notably, tumor-intrinsicglycan upregulation negatively correlates with their immune scorealignment, suggesting that tumor hyperactivation may obscure immunesignals in bulk tissue. Together, our results define a robust computationallandscape of shared metabolic alterations, providing a data-drivenframework to guide future experimental validation of candidate therapeutictargets across diverse oncological contexts.

Structural changes in protein glycosylation are recognizedas phenotypesin numerous diseases, including cancer. Despite the promise of glycoproteomicsfor diagnostic biomarker discovery, large-scale studies remain limitedby challenges in reproducibility, throughput, and quantitative precision.Here, we present a robust data-independent acquisition (DIA)-basedplasma glycoproteomics platform that enables high-throughput and reproducibleglycopeptide quantification suitable for large-cohort studies. Theworkflow integrates automated in-solution digestion and enrichment,optimized DIA-LC-MS acquisition, and a customized bioinformatics pipelineto achieve reproducible glycopeptide identification and quantitation.Across 560 replicates of a pooled human plasma sample processed inseven batches, the platform achieved high reproducibility, with intra-batchcoefficients of variation (CVs) below 15% (n = 80 per batch) and aninter-batch CV of 22.9%. The method accurately captured expected foldchanges from spiked-in glycoprotein standards. In the setting of aproof-of-concept pilot study to demonstrate the platform’sability to detect disease-associated glycosylation differences, theplatform successfully identified glycoform-specific changes and distinguishedbetween noncancer individuals (n = 39) and those with stage III/IVlung cancer (n = 39). Together, these results establish a high-throughputDIA-based glycoproteomics workflow suitable for large-cohort studiesto discover glycopeptide biomarker candidates.

One critical step in any targeted mass spectrometry experimentis selecting, from each protein of interest, a small number of peptidesthat respond well in the mass spectrometer and can serve as reliableproxies for protein quantification. Existing methods select targetpeptides either by relying on prior empirical measurements, limitingtheir applicability to previously observed peptides, or using machinelearning to predict peptide behavior from sequence alone. However,current machine learning tools suffer from various limitations, includingusing detectability as an indirect proxy for intensity, relying onsmall training sets, or ignoring the precursor charge state. In thisstudy, we introduce Bromo, a transformer-based deep learning modelthat ranks peptide precursors from a given protein by their relativeresponse, taking charge state into account. Trained on millions ofannotated peptide pairs derived from large-scale, publicly availabledata-independent acquisition mass spectrometry data, Bromo consistentlyoutperforms existing sequence-based methods across diverse, independentdata sets. Furthermore, we show that fine-tuning Bromo on experiment-specificdata can account for differences in sample preparation, sample matrix,and instrument platform, all of which influence which peptides serveas optimal targets. This adaptability makes Bromo a practical toolfor selecting target peptides for selected reaction monitoring andparallel reaction monitoring assay development across a wide rangeof experimental conditions.

Antigen interference from residual antigen peptides isa persistentchallenge in antipeptide immunocapture liquid chromatography–tandemmass spectrometry (IC-LC–MS/MS) assays, particularly when quantifyinglow-abundance protein biomarkers. Traditionally, carrier protein-conjugatedsurrogate peptides identical to those used in LC–MS/MS quantificationare employed as antigens for immunization, often resulting in varyingdegrees of antigen interference from the resulting antipeptide polyclonalantibodies (pAbs). While precapture purification and other mitigationstrategies can reduce interference, they may be insufficient for ultrasensitiveantipeptide IC-LC–MS/MS assays. Here, for the first time, wedemonstrate the use of carrier-protein-conjugated, stable isotopicallylabeled surrogate peptides as antigens for immunization. This strategyensures that residual antigen peptide is distinguishable from endogenousanalytes by mass spectrometry, effectively eliminating antigen interference.This paper details the design principles for stable isotopically labeledsurrogate peptides used in antigen preparation and compares the performanceof pAbs generated using labeled versus traditional nonlabeled peptideantigens. Results from over 20 batches of antipeptide pAbs demonstratethat this method enables the routine generation of antigen interference-freeantipeptide pAbs, facilitating the development of ultrasensitive antipeptideIC-LC–MS/MS assays for low-abundance protein biomarker quantificationwithout incurring additional time and significant cost. Data are availablevia ProteomeXchange with an identifier of PXD076272.

Metastatic gastric cancer remainsa major cause of cancer-relatedmortality, while currently available serum biomarkers have limitedperformance for metastatic risk assessment. This study aimed to identifycandidate serum biomarkers associated with gastric cancer progressionthrough quantitative proteomics and integrated multi-omics analysis.Serum data-independent acquisition (DIA) proteomics was performedin patients with metastatic gastric cancer and healthy controls, followedby integration with TCGA-STAD survival analysis and multi-algorithmmachine-learning modelling to identify clinically relevant candidates.Single-cell RNA sequencing, spatial transcriptomics, and cell–cellcommunication analyses were further applied to characterize the tumormicroenvironmental context of candidate genes. Serum ITIH3 levelswere validated by ELISA in healthy controls, non-metastatic gastriccancer, lymph-node-only metastatic gastric cancer, and distant metastaticgastric cancer groups. Proteomic analysis identified 45 differentiallyexpressed proteins, and ITIH3, VWF, SH3BGRL3, and PKLR were prioritizedthrough integrated analyses. Single-cell analyses indicated that ITIH3expression was mainly associated with cancer-associated fibroblast-relatedextracellular matrix remodeling signatures. ELISA validation demonstratedprogressively increased serum ITIH3 levels during gastric cancer progression,with favorable discrimination among selected disease groups. Thesefindings suggest that serum ITIH3 is a candidate biomarker associatedwith gastric cancer progression and tumor microenvironmental alterations,warranting further clinical validation.