Host-Defense Extracellular Vesicle Protein Changes in Antibiotic- and Staphylococcus aureus–Treated Blood

Host-Defense Extracellular Vesicle Protein Changes in Antibiotic- and Staphylococcus aureus–Treated Blood | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Host-Defense Extracellular Vesicle Protein Changes in Antibiotic- and Staphylococcus aureus–Treated Blood Dapi Menglin Chiang, Michael W. Pfaffl, Gustav Schelling, Agnes S. Meidert, and 9 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8384635/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Sepsis accounts for nearly 20% of global mortality, with antibiotic resistance worsening clinical outcomes. Rapid antibiotic administration and accurate pathogen identification remain crucial. It is well now known that extracellular vesicles (EVs) from human cells and bacterial membrane vesicles (bMVs) play a central role in the interaction between host and pathogen and represent promising biomarkers for early infections. This study investigated how antibiotic exposure alters EV responses in Staphylococcus aureus ( SA )–spiked blood and compared these findings with EV proteome profiles from bacteremia patients. Whole blood from healthy donors was spiked with SA (Multiplicity of infection: 0.001) and treated with piperacillin–tazobactam (Pip-Tazo), vancomycin, or moxifloxacin at clinically relevant concentrations. EVs were isolated using the Miltenyi Pan EV Kit, and bMVs were captured with magnetic beads conjugated to anti-OmpA and anti-GroEL. EVs and bMVs were analyzed using bead-based flow cytometry. Proteomics of total plasma, EVs and bMVs high-resolution LC–MS/MS. Patients’ serum EVs from 6 healthy controls and 12 bacteremia patients (6 blood culture–positive, 6 culture-negative) were processed using the same workflow to assess both host and bacterial proteins. Flow cytometry revealed that levels of CellMask Orange⁺ (CMO⁺) CD45⁺ PanEV⁺ SA⁺ extracellular vesicles increased in blood samples exposed to low concentrations of Pip–Tazo and high concentrations of vancomycin, despite minimal changes in vesicle size distribution and total particle counts. Proteomic analysis identified notable alterations in EV-associated proteins, including strong elevation of the ribosomal protein bL21 in SA-spiked samples and those treated with vancomycin. Gene ontology analysis indicated enrichment of innate immune and exosome-related pathways. In patient samples, EVs were enriched with acute-phase proteins such as PLSCR1, haptoglobin, CRP, and SAA1–4, along with canonical EV markers CD81 and MFGE8, irrespective of culture positivity. Antibiotic treatment leads to significant remodeling of the EV proteome, characterized by enhanced presence of immune and bacterial response proteins, even when vesicle numbers remain constant. These protein shifts appeared in both culture-positive and culture-negative patient samples, supporting the idea that EV-associated proteins could serve as early, host-derived indicators of bloodstream infection. Health sciences/Diseases Biological sciences/Immunology Biological sciences/Microbiology Figures Figure 1 Figure 2 Figure 3 Introduction Sepsis is a severe, life-threatening syndrome in which the host’s response to infection becomes dysregulated, triggering systemic inflammation, tissue injury, and the risk of multiple organ failure 1 . Sepsis continues to affect millions worldwide, with about 49 million cases and 11 million deaths each year, and the burden falls most heavily on low- and middle-income nations 1 . Sepsis commonly develops when bacteria from a localized infection enter the bloodstream. In community-onset sepsis, frequent bloodstream pathogens include Escherichia coli , Klebsiella species, Streptococcus species, and Enterococcus species 2 . Staphylococcus aureus ( SA ), including both methicillin-susceptible (MS) SA and methicillin-resistant (MR) SA strains, respectively MSSA and MRSA, is often found in blood cultures, and MRSA infections are associated with higher hospital mortality 2 . In sepsis involving SA, immune homeostasis becomes severely disrupted, with strong inflammatory activation occurring alongside an immunosuppressive state that allows the bacteria to persist and spread 3 , 4 . In early sepsis, some well known protein biomarker like S100A8/A9, pentraxin-3 (PTX3), serum amyloid A1 (SAA1), high mobility group box 1 (HMGB1), resistin, and complement C3/C5 are elevated and reflect disease severity. These key host responses including neutrophil, endothelial, and complement activation, although proteins like SAA1 and PTX3 are difficult to detect because of low baseline levels, rapid fluctuations during illness, and interference from abundant plasma proteins 3 , 5 – 7 . In contrast, clinically used markers such as C-reactive protein (CRP) are nonspecific and can give a false sense of reassurance, since CRP may remain low even in confirmed sepsis, particularly in the early phase of illness or in immunocompromised patients, which can delay timely diagnosis and treatment 8 . Together, these issues reveal the constraints of relying solely on untargeted plasma proteomics and suggest that extracellular vesicles (EVs) may provide a richer and more mechanistically relevant pool of biomarkers. EVs released by human cells, particularly small EVs (sEVs, 30–150 nm) formerly known as exosomes, carry cargo that reflects their cellular origin, and unlike low-abundance or rapidly cleared plasma proteins, they provide a concentrated and protected source of mechanistically informative mediators, making them valuable players and biomarker candidates in sepsis 9 , 10 . Bacteria also secrete vesicles of similar size, known as bacterial membrane vesicles (bMVs), which carry virulence factors and immunomodulatory molecules that can contribute to sepsis pathogenesis 11 . In sepsis, sEVs are actively released by immune cells, endothelial cells, epithelial cells, and platelets in response to infection, hypoxia, and systemic inflammation, contributing to both the propagation and regulation of host immune responses 10 , 12 . Proteome profiles of sEVs in sepsis patients often show higher levels of inflammation-related proteins, including SAA1, CRP, and immunoglobulin components, suggesting that these vesicles reflect key inflammatory processes and may aid in diagnosis 13 . Monocyte- and platelet-derived EVs in sepsis carry surface markers and cargo that promote endothelial activation, increase vascular permeability, and enhance tissue factor–mediated coagulation, directly contributing to microvascular dysfunction and multi-organ failure 14 . Although interest in sEVs as biomarkers in sepsis is growing, the dynamic response of the host sEVs proteome to direct bacterial challenges, such as SA infection, remains poorly understood. Furthermore, how antibiotic treatment shapes this response remains unclear because comprehensive comparisons among culture-positive sepsis, culture-negative sepsis, and healthy individuals are limited, even though such analyses are essential for defining diagnostic specificity and clinical usefulness. In this study, we investigated the human EV proteomic landscape using a dual approach: (1) an ex vivo whole-blood infection model challenged with SA , with or without antibiotic treatment, and (2) clinical plasma samples from patients with blood culture–positive and –negative sepsis, alongside healthy volunteers. Using immunomagnetic bead–based EV isolation together with label-free proteomics and multivariate analysis, we set out to define EV signatures that track with infection status and treatment effects. This approach allowed us to detect both human and bacterial proteins within the vesicles and to observe how antibiotic exposure influenced their composition. Building on these findings, we compared EV proteomes from culture-positive sepsis, culture-negative sepsis, and healthy individuals to define EV signatures that reflect infection status and treatment response in sepsis. Methods SA-Spiked Plasma Sample Collection Peripheral blood (55 ml) was collected from six healthy volunteers in EDTA-coated tubes to prevent coagulation. From each donor, 5 ml of whole blood was distributed into 15 ml Falcon tubes and allocated to 11 experimental conditions: a MOCK (non-infection) control, infection with SA DSM 20231 (MSSA), and infection in combination with nine antibiotic treatments. The SA DSM 20231 strain was obtained from DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen, German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany) and maintained on tryptic soy broth (TSB) agar plates. A single colony was inoculated into 500 ml of RPMI-1640 medium, prepared as two 250-ml cultures, and grown at 37°C, yielding a bacterial suspension with an OD₆₀₀ of approximately 1.0–1.5. From the resulting culture mixture, a 1-ml aliquot was labeled with the CFSE Cell Division Tracker Kit (BioLegend, San Diego, USA) for 30 min at 37°C. CFSE-labeled SA were quantified by flow cytometry (LSRFortessa™, Becton, Dickinson and Company, Franklin Lakes, NJ, USA), and data analysis was performed using FlowJo 10.8 (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). For infection, 3 × 10⁷ SA were added to 5 ml of whole blood, corresponding to a multiplicity of infection (MOI) of 0.001 and approximately 6 × 10⁹ cells/ml as measured with a TC10™ Automated Cell Counter (Bio-Rad). Antibiotic concentrations were selected to represent clinically relevant plasma levels: piperacillin–tazobactam (25, 50, 100 µg/ml), vancomycin (6.25, 12.5, 25 µg/ml), and moxifloxacin (0.18125, 0.725, 2.9 µg/ml). These antibiotics were selected for their distinct mechanisms of action, including inhibition of cell wall synthesis by piperacillin–tazobactam (Pip-Tazo) and vancomycin, and inhibition of DNA replication by moxifloxacin, in accordance with the Surviving Sepsis Campaign 2021 international guidelines for empiric treatment of SA and sepsis 15 . All 11 experimental conditions were used for flow cytometric analysis, whereas only the highest concentration of each of the three antibiotics (100 µg/ml Pip-Tazo, 25 µg/ml vancomycin, and 2.9 µg/ml moxifloxacin), together with the MOCK control and infection-alone condition, was used for subsequent proteomic analyses. Following 24 h incubation at 37°C, plasma was separated by centrifugation at 2,000 × g for 10 min at 4°C and filtered through a 0.45 µm PES syringe filter (Merck KGaA, Darmstadt, Germany) to remove residual bacteria. Approximately 2 ml of filtered plasma was collected per condition and stored at − 80°C until analysis. Human EV isolation Human EVs were isolated from (i) 1 ml of filtered plasma obtained from the SA-spiked blood model (MOCK control, infection-only, and antibiotic-treated conditions; n = 6 per condition) and (ii) serum EV–Miltenyi bead complexes collected from healthy controls (HC, n = 6), bacterial culture–negative patients (BCN, n = 6), and bacterial culture–positive patients (BCP, n = 6). EV isolation was performed using immunoaffinity capture with the EV Isolation Kit Pan, human (Miltenyi Biotec, Bergisch Gladbach, Germany), following the manufacturer’s protocol. Briefly, 1 ml of filtered plasma or serum was incubated with 50 µl of immunoaffinity beads for 1 hour on an overhead rotor. A Miltenyi µColumn was equilibrated and washed three times before sample loading, followed by four additional washes on a magnetic stand. EV–Miltenyi bead complexes were eluted using 100 µl of the kit-supplied isolation buffer. The unbound plasma fraction (1 ml) collected during the immunoaffinity procedure was retained for additional downstream analyses. Clinical characteristics of all participants, including age, sex, hospital and intensive care stay, pathogen distribution, day 1 antibiotic treatment status, and day 1 inflammatory markers (CRP, interleukin-6, leukocyte count, and procalcitonin), are summarized in Table 1 . Table 1 Summary of clinical characteristics. Characteristics BCP (n = 6) BCN (n = 6) Age (years) – Mean ± SD 67.8 ± 16.7 74.8 ± 15.7 Male 4 3 Female 2 3 Hospital stay – Mean ± SD 3.8 ± 2.8 9.3 ± 6.8 Intensive care – Mean ± SD 3.0 ± 2.7 3.7 ± 3.2 Pathogens SA (4), S. capitis (1), C. tertium (1) None Day 1 antibiotic treatment – Yes 5 3 Day 1 antibiotic treatment – No 1 3 Day 1 C-reactive protein 22.7 ± 16.3 22.4 ± 14.5 Day 1 Interleukin-6 567.9 ± 590.2 199.8 ± 203.5 Day 1 Leukocytes count 11.1 ± 7.1 20.2 ± 13.8 Day 1 Procalcitonin 13.9 ± 12.4 2.9 ± 3.1 In-house bMV isolation The bacterial beads (BacB) necessary for the bMVs isolation is described in brief: First, 250 µl of 40 µg/ml 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Thermo Fisher Scientific, Bremen, Germany) in 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer (Sigma + Merck KGaA, Darmstadt, Germany), pH 5.0, was used to activate 10 mg of carboxyl-group magnetic particles (2.13 × 10¹⁰ particles/ml, IKERLAT Polymers, Lasarte-Oria, Spain) at 25°C for 15 minutes. The activated magnetic particles were then resuspended in 1 ml of PBS. Next, 50 µg each of OmpA antibodies (catalog number: orb26813, Biorbyt, Cambridge, UK) and GroEL antibodies (catalog number: NBP2-32867, Novus Biologicals, Centennial, United States) were prepared. A total of 100 µl of the antibody mixture was incubated with 1 ml of activated magnetic particles at 37°C for 2 hours. To block unoccupied binding sites, the antibody-coated beads were incubated with 1 ml of 10% EV-free BSA in PBS (prepared by centrifugation at 100,000 × g, 4°C, overnight) at 37°C for 30 minutes. The resulting beads were then diluted to 1 × 10⁶ beads/µl and used as BacB. One milliliter of unbound plasma was incubated with 100 µl of BacB at 37°C for 1 hour. After incubation, bMV–BacB complexes were washed twice with 1 ml of 0.5% EV-free BSA in PBS while magnetic bMV–BacB complexes were retained in the tube using a magnetic stand. As a negative control, 1 ml of 10% EV-free (EF) BSA in PBS was incubated with BacB under the same conditions to generate EF–BacB complexes. bMV–BacB complexes were used for further analysis by bead-based flow cytometry or proteomics, while EF–BacB complexes were used only as a negative control for bead-based flow cytometry. Bead-based flow cytometry 25 µl of EV–Miltenyi bead complexes or 250 µl of bMV–BacB complexes were diluted in 0.9 ml of 0.5% EV-free BSA in PBS. To this mixture, 100 µl of an antibody master mix was added. The master mix contained 2.5 µl of PanEV antibody (3 µg), pre-mixed with 1 µl each of PE-conjugated anti-CD63 (clone H5C6), PE-conjugated anti-CD81 (clone 5A6), and PE-conjugated anti-CD9 (clone HI9a) (all from BioLegend, San Diego, USA); 5 µl of Brilliant Violet 421™ anti-human CD45 antibody (clone HI30, BioLegend, San Diego, USA); 10 µM CellMask™ Plasma Membrane Deep Red (CMD; C10046, Thermo Fisher Scientific, Bremen, Germany); and 1 µl of FITC-conjugated SA polyclonal antibody (PA1-73172, Thermo Fisher Scientific, Bremen, Germany). The mixture was incubated at 37°C for 1 hour. After staining, EV–Miltenyi bead complexes and bMV–BacB complexes were washed twice with 1 ml of 0.5% EV-free BSA in PBS and centrifuged at 100,000 × g at 4°C overnight using a magnetic rack. The samples were then analyzed on a BD LSR Fortessa™ flow cytometer (Becton, Dickinson and Company, Franklin Lakes, USA). Data analysis, including gating, was performed using FlowJo software version 10.8 (Becton, Dickinson and Company, Franklin Lakes, USA). An IgG isotype control was included to confirm the specificity of the staining. In the flow cytometry gating strategy, EV–Miltenyi bead complexes and bMV–BacB complexes were initially gated based on forward scatter (FSC-A) and side scatter (SSC-A). The CMD⁺ PanEV⁺ CD45⁺ Anti-SA⁺ population from the EV–Miltenyi bead complexes was identified as immune cell-derived EVs carrying SA antigen. In contrast, the CMD⁺ PanEV⁻ Anti-SA⁺ population from the bMV–BacB complexes was identified based on comparison to the isotype control. Human EV characterization Nanoparticle tracking analysis (NTA) was used to determine the size distribution and particle concentration of extracellular vesicles (EVs) using ZetaView® PMX 110 or x30 MONO instruments equipped with a 520-nm laser and a 550-nm emission filter (Particle Metrix GmbH, Inning am Ammersee, Germany). EV samples were diluted in particle-free PBS to fall within the optimal measurement range of the instrument and were measured in two acquisition cycles across 11 positions per sample. For fluorescence measurements, EV membranes were labeled with CellMask Orange™ (CMO; 5 mg/mL stock; Thermo Fisher Scientific Inc., Waltham, MA, USA) at a final concentration of 25 µg/mL (1:200 dilution), and samples were further diluted 1:400 prior to analysis unless otherwise stated. Fluorescence measurements on the PMX 110 system were performed using 1 mL of filtered plasma obtained from a Staphylococcus aureus –spiked blood model (mock control, infection-only, and antibiotic-treated conditions; n = 6 per group) with video capture settings of shutter speed 70, sensitivity 95, and a frame rate of 30 frames/s, and data were analyzed using ZetaView software version 8.05.11 SP1. Fluorescence measurements on the x30 MONO system were conducted on serum EV–Miltenyi bead complexes from healthy controls (HC; n = 6), bacterial culture-negative patients (BCN; n = 6), and bacterial culture-positive patients (BCP; n = 6) using a shutter speed of 100, sensitivity 90, and a frame rate of 30 frames/s, with data analyzed using ZetaView software version 8.06.01 SP1 (Particle Metrix GmbH, Inning am Ammersee, Germany). Cryo-electron microscopy (cryo-EM) sample preparation was conducted as previously described 16 . In brief, 4 µl of each sample was applied to glow-discharged holey carbon-coated grids (Ted Pella Inc., Redding, United States), blotted using Whatman No. 1 filter paper, and vitrified by plunge-freezing into liquid ethane at − 178°C with a Leica GP plunger (Leica Microsystems GmbH, Wetzlar, Germany). Vitrified grids were transferred to a Talos electron microscope (FEI/Thermo Fisher Scientific Inc., Waltham, United States) using a Gatan 626 cryo-holder (Gatan Inc., Pleasanton, United States). Imaging was performed at an accelerating voltage of 200 kV under low-dose conditions (20 e⁻/Ų) while maintaining cryogenic temperatures. Micrographs were acquired using a CETA camera (Thermo Fisher Scientific Inc., Waltham, United States). Negative staining transmission electron microscopy (TEM) was performed as previously described 16 . Briefly, 5 µl of undiluted sample was adsorbed for 60 seconds onto glow-discharged parlodion/carbon-coated copper grids (FCF200-Ni; Electron Microscopy Sciences, Hatfield, United States). The grids were then blotted, rinsed three times with double-distilled water (ddH₂O), and stained on two consecutive droplets of 2% uranyl acetate solution (Merck KGaA, Darmstadt, Germany). Imaging was carried out using a Talos F200C TEM (FEI/Thermo Fisher Scientific Inc., Waltham, United States) operated at an accelerating voltage of 120 kV. Electron micrographs were acquired using a Veleta camera (EMSIS GmbH, Münster, Germany). Protein preparation for proteomics EV proteins from (i) plasma EV–Miltenyi bead complexes, unbound plasma bMV–BacB complexes, and total plasma obtained from the SA-spiked blood model (MOCK control, infection-only, and antibiotic-treated conditions; n = 6 per condition), and from (ii) serum EV–Miltenyi bead complexes collected from healthy controls (HC, n = 6), bacterial culture–negative (BCN, n = 6), and bacterial culture–positive (BCP, n = 6) patient samples were lysed in 1× RIPA buffer (Abcam plc, Cambridge, United Kingdom; ab156034) supplemented with 1× ProteaseArrest™ Protease Inhibitor Cocktail (G-BIOSCIENCES, St. Louis, MO, United States), and stored at − 80°C until further use. Protein samples were then boiled at 70°C for 10 minutes and sonicated on ice for 5 minutes. Following sonication, lysates were centrifuged at 10,000 × g at 4°C for 30 minutes. Protein concentrations were determined using the bicinchoninic acid (BCA) assay (Thermo Fisher Scientific Inc., Waltham, United States). A minimum of 5 µg of protein from each sample was mixed with 1× Laemmli Sample Buffer (Bio-Rad Laboratories, Hercules, United States) containing 2-mercaptoethanol (Merck KGaA, Darmstadt, Germany) in a final volume of 35 µl. Prepared samples were then submitted to the proteomics core facility at the Bavarian Center for Biomolecular Mass Spectrometry for analysis. Mass spectrometry-based proteomics In accordance with standard procedures, in-gel trypsin digestion was performed 17 on all samples derived from both experimental datasets. This included (i) EV–Miltenyi bead complexes, bMV–BacB complexes, and total plasma obtained from the SA-spiked blood model (MOCK control, infection-only, and antibiotic-treated conditions; n = 6 per condition), as well as (ii) EVs and total plasma or serum collected from healthy controls (HC, n = 6), bacterial culture–negative (BCN, n = 6), and bacterial culture–positive (BCP, n = 6) patient samples. Briefly, each sample was loaded onto a NuPAGE™ 4–12% Bis-Tris protein gel (Thermo Fisher Scientific, Waltham, USA) and run approximately 1 cm into the gel to concentrate the proteins into a single non–size-separated band. The accumulated band was excised, reduced with 50 mM dithiothreitol, alkylated with 55 mM chloroacetamide, and digested overnight at 37°C with Trypsin Gold (mass spectrometry grade, Promega). Following digestion, peptides were extracted, dried, and resuspended in 25 µl of buffer A (2% acetonitrile, 0.1% formic acid in HPLC-grade water). A 5 µl aliquot of each peptide solution was injected for LC–MS/MS analysis. LC-MS/MS measurements were conducted using a Dionex Ultimate 3000 RSLCnano system linked to an Orbitrap Fusion LUMOS instrument (ThermoFisher Scientific, Bremen), in adherence to the standard protocol of the core facility 18 , 19 . Samples were initially loaded onto a self-packed trap column (ReproSil-pur C18-AQ, 5 µm, 20 mm × 75 µm, Dr. Maisch) at a flow rate of 5 µL/min using 0.1% formic acid. After a 10-minute loading step, peptides were eluted onto a self-packed analytical column (ReproSil Gold C18-AQ, 3 µm, 450 mm × 75 µm, Dr. Maisch). Chromatographic separation occurred over a 50-minute gradient increasing from 4% to 32% solvent B (acetonitrile containing 0.1% FA and 5% DMSO) against solvent A (water containing 0.1% FA and 5% DMSO) at a flow rate of 300 nL/min. The mass spectrometer operated in positive mode using data-dependent acquisition (DDA). We acquired full MS1 scans (360–1300 m/z) at a resolution of 60,000, with a normalized AGC target of 100% and a maximum injection time of 50 ms. A 2-second cycle time was employed, selecting precursors with charge states between 2 and 6, while applying a 30-second dynamic exclusion. Fragmentation was carried out via higher-energy collision-induced dissociation (HCD) at 30% normalized collision energy (NCE) with a 1.3 m/z isolation window. MS2 spectra were recorded at a resolution of 15,000, using an AGC target of 150% and a maximum injection time of 22 ms. Label-Free Quantitative Proteomics and Integrated Statistical Analysis for Host–Pathogen Profiling in Sepsis Proteomic peptide identification and quantification were executed in accordance with our previous study 19 , utilizing the MaxQuant computational platform (version 1.6.3.4) and the integrated Andromeda search algorithm 20 . Tandem mass spectrometry (MS/MS) spectra were queried against UniProt reference proteomes for Homo sapiens (Taxon ID: 9606, UP000005640, July 2020) and Staphylococcus aureus (Taxon ID: 1280, July 2024), with the search space augmented by the standard MaxQuant contaminant database. Search parameters were configured with Trypsin/P specificity; carbamidomethylation of cysteine was defined as a fixed modification, while methionine oxidation and N-terminal acetylation were treated as variable modifications. To ensure statistical reliability, a target-decoy strategy utilizing reversed sequences was applied, restricting the false discovery rate (FDR) to 1% at both the peptide spectrum match (PSM) and protein levels. Subsequent data curation was conducted in Perseus (version v2.0.10.0) 21 . The dataset was filtered to exclude proteins identified merely by site, reverse hits, and potential contaminants, after which label-free quantification (LFQ) intensities underwent log10 transformation 21 . To identify proteins reproducibly detected across sample types including EV–Miltenyi bead complexes, bMV–BacB complexes, total plasma, and patient serum, proteins were retained if they had valid values in at least 50% of biological replicates within any one group, using the “Filter rows based on valid values” function in Perseus 21 . Missing values were imputed using the standard normal distribution–based imputation in Perseus, followed by quantile normalization using the built-in Perseus normalization function. These preprocessing steps were applied consistently to both datasets: (i) the SA –spiked plasma experiment and (ii) the patient cohort comprising HC, BCN, and BCP groups. Preprocessed data were subsequently analyzed in RStudio (version: 2025.05.0 + 496) 22 with R (version: 4.5.1) 23 . Differential expression analysis was performed separately for the two datasets. For the SA –spiked plasma experiment (i), paired two-sided t-tests were used to account for matched biological replicates from the same donors across conditions. For the patient cohort (ii), unpaired two-sided Student’s t-tests were applied, reflecting the independence of clinical samples. In both analyses, proteins were considered significantly regulated if they exhibited p < 0.05 and an absolute log₁₀ fold change (|log₁₀FC|) greater than 0.301, corresponding to a fold change greater than 2 or less than 0.5. Volcano plots were generated using ggplot2, and significant proteins were annotated using a combined human and SA protein database. Protein-level functional enrichment analysis was performed using the clusterProfiler (version: 4.14.6) 24–27 , ReactomePA (version: 1.50.0) 28 , and org.Hs.eg.db (version: 3.20.0) packages to identify enriched Gene Ontology (GO) terms, KEGG pathways, and Reactome pathways. Enrichment significance was evaluated using Fisher’s exact test. Results were visualized via pathway–protein heatmaps, filtered using sepsis-related biological keywords. To assess global expression trends and group separation, sparse Partial Least Squares Discriminant Analysis (sPLS-DA) was performed using the mixOmics package (version: 6.30.0) 29 . Component-based plots with group-specific clustering and confidence ellipses were generated in RStudio (version 2025.05.0 + 496) 22 with R (version 4.5.1) 23 . Finally, Venn diagrams were constructed using the Venn Diagram package to visualize the overlap and uniqueness of differentially expressed proteins across clinical conditions. Results To assess the impact of SA infection, we isolated EVs and bMVs from plasma using two approaches: Miltenyi beads for human EV capture and BacB-coated magnetic beads for bMV–EV complex enrichment (Fig. 1 A). Fluorescence nanoparticle tracking analysis (F-NTA) and negative-stain electron microscopy of the EV–Miltenyi bead complexes first showed no observable differences across all experimental groups (Fig. 1 B–D). Negative-stain electron microscopy of bMV-BacB complexes also showed no observable differences across all the groups (Supplemental Fig. 1A-E). Analysis of the EV–Miltenyi bead complexes revealed a significant increase in CMD⁺ PanEV⁺ CD45⁺ SA + signals in the SA-spiked group compared to the MOCK control ( p = 0.032) (Fig. 1 E). Within the SA -spiked groups, treatment with either the lowest concentration of piperacillin-tazobactam (25 µg/ml) ( p = 0.032) or the highest concentration of vancomycin (25 µg/ml) ( p = 0.032) also resulted in significantly higher CMD⁺ PanEV⁺ CD45⁺ SA⁺ signals compared to the SA-spiked group without antibiotics (Fig. 1 E). Analysis of the bMV fraction, using unbound plasma collected after Miltenyi bead isolation, revealed that the bMV–BacB complexes exhibited a significant increase in CMD⁺ PanEV⁺ SA⁺ signals in the SA -spiked group compared to the MOCK control ( p = 0.032) (Fig. 1 F). However, no significant differences were observed between the antibiotic-treated groups and the SA -spiked group without treatment (Fig. 1 F). These findings suggest that SA infection induces significant changes in the protein composition of EVs and bMVs, rather than altering their total quantity, as indicated by EV flow cytometry and F-NTA. To better understand how SA infection with or without antibiotic treatment alters the protein composition of EVs and bMVs, we carried out proteomic profiling on both fractions using the antibiotic concentrations that produced the strongest effects in our flow cytometry assays. We also included total plasma proteomics to provide an overall view of systemic protein changes. However, the protein yields from the bMV fraction and from total plasma were too low for reliable pairwise comparisons (Supplemental Fig. 2A–B). For this reason, we focused all downstream analyses on the EV fraction. sPLS-DA plotting showed clear separation between the MOCK control and the SA-infected/antibiotic-treated groups (Fig. 2 A). Furthermore, DEFA3, LTF, SA ribosomal protein rplU, ITGB2, RAB7A, S100A8, UBA52, PTPRC, TF, ITGAM, SERPINA1, MYADM, SERPINC1, IQGAP2, ORM1, A1BG, SDCBP, C1QB, HPX, and TNFAIP8 contributed most to the variance in component 1 (Fig. 2 B). A volcano plot showed that DEFA3, LTF, rplU, MYADM, SDCBP, ITGB2, UBA52, PTPRC, S100A8, S100A9, ITGAM, RAB7A, and ITGA6 were significantly upregulated in the SA infection group compared to the MOCK control (Fig. 2 C). A Venn diagram demonstrated that DEFA3, LTF, ITGB2, RAB7A, S100A8, UBA52, PTPRC, ITGAM, MYADM, and SDCBP were commonly identified in both the sPLS-DA and volcano plot analyses (Fig. 2 D). Enrichment analysis revealed that these 11 proteins were significantly associated with pathways related to positive regulation of secretion, EV biogenesis, macrophage activation, bacterial infection, and host defense responses (Fig. 2 E). These proteomic results further support our EV flow cytometry findings (Fig. 1 B). However, pairwise comparisons did not identify any significantly upregulated proteins in the piperacillin-tazobactam group compared to the SA-only group. In contrast, AZU1, EPB41, CAMP, and IGKV2-30 were significantly downregulated in the piperacillin-tazobactam group (Fig. 2 F). Interestingly, SA proteins RplU and InfC were significantly increased in the vancomycin-treated group relative to the SA-only group (Fig. 2 G), supporting the EV flow cytometry data that indicated enhanced loading of SA -derived proteins into human EVs following vancomycin treatment (Fig. 1 B). In the comparison between the moxifloxacin group and the SA -only group, C1QB, PRKACB, SH3BGRL3, and OAZ1 were upregulated, whereas AZU1, APOD, IGKV2-30, and FCN2 were downregulated. Enrichment analysis did not reveal any significantly associated human pathways, consistent with the broadly similar proteomic profiles observed across the three antibiotic treatment groups compared to SA infection alone. Since EV–Miltenyi bead flow cytometry and EV proteomics yielded the most reproducible results in SA –spiked plasma, we next investigated whether the human proteome is similarly altered in sepsis patients and whether bacterial ribosomal proteins can be detected in EVs. Prior to proteomic analysis, EV particle size and number were consistent across HC, BCN, and BCP groups (Supplemental Fig. 3A and B). Following data preprocessing steps including quality control, intensity normalization, and reproducibility-based filtering across biological replicates, a total of 424 proteins were confidently identified across samples from HC ( n = 6), BCN ( n = 6), and BCP ( n = 6) groups. sPLS-DA analysis revealed clear separation between the HC group and the infection groups (Fig. 3 A). Notably, the overall proteomic profile of BCP patients also showed distinct separation from that of BCN patients, indicating further stratification based on blood culture status (Fig. 3 A). Differential analysis identified 57 significantly enriched proteins in the BCN group and 69 in the HC group (fold change > 2) (Fig. 3 B). Among the top five uniquely enriched proteins in BCN patients were SAA1, ANXA1, PLSCR1, SERPINA3, and APCS. Pathway enrichment analysis of the BCN-associated proteins revealed significant involvement in blood microparticles, positive regulation of exocytosis, defense response to bacteria, and immune-related pathways (Fig. 3 C). In the BCP group, 36 proteins were significantly enriched compared to HC, while 38 were enriched in HC (fold change > 2) (Fig. 3 D). The top five uniquely enriched proteins in BCP patients were CRP, SAA2, SERPINA1, SERPINA3, and ANXA2. Enrichment analysis of BCP-associated proteins revealed similar involvement in blood microparticles, exocytosis, bacterial defense, and immune pathways (Fig. 3 E). Notably, the SA infection pathway was also significantly enriched in BCP patients relative to HC (Fig. 3 E). The observed enrichment of pathways related to blood microparticles and exocytosis in both BCN and BCP patients, as revealed by proteomic analysis, further supports our EV flow cytometry findings (Fig. 1 B). To further explore host responses in culture-stratified infection states, we compared the proteomic profiles of BCN and BCP patients directly. Ten proteins were significantly enriched in each group (Supplemental Fig. 3C). Proteins enriched in BCN patients, including the metabolic enzyme LDHA; immune modulators ANXA1, ANXA3, ANXA5, and ANXA6; the oxidative burst enzyme CYBB; inflammatory mediator S100A9; complement-associated CLU; eosinophil granule protein PRG2; and antimicrobial peptides DEFA1/DEFA3, were associated with bacterial defense responses, endocytic vesicle formation, and leukocyte chemotaxis (Supplemental Fig. 3D). In contrast, proteins enriched in BCP patients, including immunoglobulin variants, ARHGDIB, NAPA, UNC13D, DSTN, HBB, and ESAM, were linked to immune recognition, vesicle trafficking, cytoskeletal remodeling, and vascular function, highlighting distinct host responses in blood culture–positive sepsis (Supplemental Fig. 3C). Additionally, 24 proteins were commonly enriched in both BCN and BCP groups, based on a Venn diagram analysis of significant proteins across the three pairwise group comparisons (Fig. 3 F). To assess global proteomic changes associated with infection, we compared the combined infection group (Bac_infection_group: BCN + BCP, n = 12) to HC (n = 6). This analysis identified 57 proteins significantly enriched in the infection group and 70 enriched in HC (Fig. 3 G). Enrichment analysis of the infection-associated proteins again highlighted key pathways including blood microparticles, positive regulation of exocytosis, SA infection, secretory granule membrane, innate immune activation, complement activation, and chemotaxis (Fig. 3 H). Together, these findings indicate that EV proteomes from sepsis patients, whether blood culture–positive or –negative, are characterized by enhanced activity in vesicle-mediated transport, innate immune activation, and host defense pathways when compared to HC. Discussion This study presents a multi-tiered investigation into how EVs reflect host responses to bacterial infection and antibiotic treatment, combining an in vitro SA infection model with EV profiling from patients with culture-positive and culture-negative sepsis. Using a combination of EV flow cytometry, bead-based isolation, and high-resolution label-free proteomics, we demonstrate that the EV proteome dynamically changes in response to the bacterial challenge and that these changes are modulated by specific antibiotic treatments. Notably, we identified distinct proteomic signatures in culture-positive and culture-negative sepsis patients that share common immune activation pathways but also exhibit stratifying features based on blood culture status. Our ex vivo data demonstrated that SA infection leads to a significant increase in CD45⁺ host-derived EVs carrying bacterial antigens (Fig. 1 B), indicating active vesicle-mediated host–pathogen interaction. This signal was further amplified by vancomycin treatment, which correlated with increased detection of SA ribosomal proteins such as RplU and InfC in the EV fraction (Figs. 1 B and 2 G). These results suggest that vancomycin may facilitate the incorporation of bacterial components into host EVs, possibly through enhanced bacterial lysis or altered vesicle trafficking pathways. Notably, fluorescence nanoparticle tracking analysis (F-NTA) showed no significant differences in the total number of EVs across experimental groups (Figs. 1 D–F), indicating that changes in vesicle content, rather than vesicle abundance, reflect the host response to infection and antibiotic treatment. In our SA model, vancomycin and piperacillin-tazobactam, which target the bacterial cell wall and induce membrane stress, enhanced the incorporation of bacterial proteins into host EVs, whereas moxifloxacin, which acts on intracellular enzymes without damaging the membrane, showed no such effect, highlighting the importance of cell wall disruption in driving vesicle-associated protein release. This observation aligns with growing evidence that bactericidal antibiotics, including β-lactams and glycopeptides, can stimulate the release of bMVs as part of a general stress response 30 . These vesicles often carry cytoplasmic and periplasmic proteins, nucleic acids, and misfolded proteins, contributing to bacterial survival, antibiotic resistance, and immune modulation 30 . In gram-negative bacteria, early electron microscopy studies demonstrated that polymyxin treatment in Pseudomonas aeruginosa and Escherichia coli ( E.coli ) markedly increased membrane vesicle production 31 . Moreover, bMVs protect E. coli and enterotoxigenic E. coli ( ETEC ) from polymyxin B, colistin, and T4 bacteriophage by adsorbing these agents. Treatment with these antibiotics induces a substantial increase in bMV output, as measured by bMV protein content normalized to the number of CFU (colony forming unit) 32 . Although SA lacks an outer membrane, our findings suggest that similar stress-responsive vesicle dynamics may occur in this gram-positive context, particularly in response to bactericidal antibiotic treatment. We used the DSM 20231 strain, a type strain from clonal complex 8 that carries multiple methicillin resistance genes, reflecting an intrinsic potential for β-lactam resistance. Consistent with prior work by Kim et al. 33 , we observed that vancomycin treatment enhanced the incorporation of bacterial ribosomal proteins such as RplU and InfC into host-derived EVs, likely due to increased bacterial lysis or altered processing of bacterial components. However, our study differs from Kim’s in two important ways. First, whereas Kim et al. 33 used pure bacterial cultures and directly isolated bMVs, our approach involved spiking SA into human plasma and isolating host-derived EVs for proteomic analysis. Second, while Kim’s study identified bacterial β-lactamases and penicillin-binding proteins within MRSA-derived EVs that conferred antibiotic protection, we did not detect such resistance enzymes in the investigated EVs. This difference likely reflects the dominance of host proteins in our plasma-based samples, which can obscure lower-abundance bacterial proteins. Nevertheless, we were able to detect SA ribosomal proteins within host EVs: RplU was significantly elevated in the SA –only group compared to the MOCK control, and both RplU and InfC showed further increases following vancomycin treatment. These findings indicate that bacterial components are successfully incorporated into host EVs in response to infection and antibiotic stress. Similar effects were also observed with ampicillin, a β-lactam that, like vancomycin, targets the peptidoglycan layer and induces membrane destabilization 30 , 34 . Although their molecular targets differ, both antibiotics appear to promote bacterial protein release through cell wall damage, facilitating their entry into host EVs. These observations are in line with Andreoni et al. 35 , who showed that certain β-lactam antibiotics such as flucloxacillin and ceftaroline enhanced vesicle formation in SA through peptidoglycan weakening, and with Kim et al. 36 , who demonstrated that sub-inhibitory concentrations of antibiotics triggered MV release in Enterococcus faecium and enhanced pro-inflammatory host responses. Together, these results support a broader role for bactericidal antibiotics in shaping EV content during host–pathogen interaction. In line with growing evidence that EV reflect immune dynamics in systemic infection, our study demonstrates that both BCP and BCN sepsis patients exhibit robust changes in the EV proteome compared to HC. Notably, we observed substantial enrichment of innate immune mediators, complement components, and acute-phase proteins in plasma EVs, consistent with prior findings summarized in Chang Tian et al. 37 and other studies. Notably, several proteins known to mediate inflammation, immune cell recruitment, and opsonization were elevated, including CRP, SAA1, SAA2, SAA4, SERPINA1, SERPINA3, APCS, ANXA1, and CD177. Complement-related proteins such as C1QA, C1R, C1S, and SERPING1 were also enriched, along with coagulation and vascular injury markers like FGG, VWF, and ITIH3. Additional upregulated proteins such as PLSCR1, SDCBP, HP, ORM1, AGT, CTSG, MFGE8, CP, and SLC25A3 further support active EV-mediated immune signaling and endothelial responses (Fig. 3 ). These protein signatures reinforce the concept that EVs act as carriers of immune and inflammatory mediators in sepsis, offering insight into both shared and context-specific pathways of disease progression. CRP, a known common sepsis marker 38 , 39 was also detected in our EV proteome, consistent with findings by Yan Xu et al. 39 . However, we did not detect Serine palmitoyltransferase 3 (SPTLC3), possibly due to differences in EV isolation methods, as their study used ultracentrifugation with density gradients, while we employed antibody bead–based capture. Chanhee Park et al. 40 and Murao et al. 13 both identified key EV proteins linked to sepsis, several of which overlap with our findings. These include CRP, SERPIN (A1, A3, and G1), ORM1, VMW, SAA (1, 2, and 4), complements (C1R, C1QA, and C1S), and ITIH3, all of which were enriched in the sepsis EV proteome of this study. Notably, SERPINA1 and SERPINA3 are inhibitors of neutrophil serine proteases, playing dual roles in controlling inflammation during sepsis 40 , 41 . In parallel, our identification of C1QA, C1R, C1S, and SERPING1 underscores activation of the classical complement pathway, reinforcing its critical role in driving proinflammatory and prothrombotic responses in sepsis. These complement components contribute not only to pathogen recognition and clearance but also to endothelial activation, immune cell modulation, and the regulation of T-cell responses, further supporting the central involvement of the complement system in EV-mediated immune dysregulation during sepsis 40 , 42 – 44 . Interestingly, several proteins identified in our dataset, such as SDCBP (syntenin), PLSCR1, CP, HP, AGT, FGG, APCS, CTSG, SLC25A3, MFGE8, and CD177, were not reported in their analyses. This suggests that our antibody bead-based EV isolation method may reveal additional or less-characterized components of the host response in sepsis. Notably, SDCBP (syntenin) was overexpressed in both the BCN and BCP groups compared to HC (Fig. 3 ). This observation is consistent with our flow cytometry analysis using EVs isolated by Miltenyi beads (Fig. 1 ), as well as with our previous publication 45 , where western blot analysis also showed an increasing trend of SDCBP expression in sepsis patients relative to healthy controls. However, despite these changes in protein expression, the particle numbers remained unchanged in both our previous study and the current dataset (Fig. 1 ). Together, these findings suggest that sepsis primarily alters the molecular composition of EVs rather than their overall abundance. Our findings support that PLSCR1, which is overexpressed in sepsis patients compared to HC, plays a key role in protecting lung epithelial cells from SA α-toxin–induced cell death by mediating IFNα-driven resistance mechanisms 46 . Ceruloplasmin (CP), a classic acute-phase protein, has recently been implicated in sepsis-induced cardiotoxicity through its involvement in cuproptosis, a newly identified form of regulated cell death 47 . Haptoglobin (HP) is another well-established acute-phase protein that mitigates hemoglobin-driven oxidative stress 48 . Angiotensinogen (AGT) shows a variable but inducible response to inflammatory stimuli, suggesting context-dependent regulation during acute stress 49 . Fibrinogen gamma chain (FGG), particularly its γ′ variant, contributes to host defense in SA sepsis by modulating bacterial aggregation and enhancing survival outcomes 50 . SLC25A3, a mitochondrial carrier involved in transporting solutes like citrate and succinate, is associated with metabolic reprogramming during inflammation, although its specific role in sepsis remains unclear 51 . MFGE8, previously reported to be downregulated in septic patients and protective against oxidative stress and ferroptosis, was found to be elevated in EVs from the BCN and BCP groups in our study; however, the functional significance of EV-associated MFGE8 in sepsis remains to be elucidated 52 . CD177, a neutrophil activation marker involved in chemotaxis, was significantly upregulated in BCN and BCP compared to healthy controls (Fig. 3 ), consistent with previous studies showing elevated CD177 expression in circulating neutrophils during septic shock 53 and increased levels of CD177⁺ neutrophil-derived microvesicles in patients with bacteremia 54 . In conclusion, our study validates many established immune-modulating components of EVs described in the literature while adding novel evidence for underappreciated proteins in both culture-positive and culture-negative sepsis patients. The enrichment of antimicrobial peptides, annexins, Reactive oxygen species-generating enzymes, and vesicle trafficking proteins in patient-derived EVs expands the current understanding of EV-mediated immune regulation. These findings support the utility of EV profiling as a tool for uncovering mechanistic differences between sepsis subtypes and for identifying candidate biomarkers relevant to early diagnosis, pathogen-independent host response, and targeted therapy. Declarations Code availability The R scripts used for nanoparticle tracking analysis (NTA), extracellular vesicle (EV) flow cytometry, and proteomics data analysis, together with the imputed and normalized ProteinGroups files generated by MaxQuant and Perseus, have been uploaded to GitHub. https://doi.org/10.5281/zenodo.17961832 Ethics statement The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Faculty of Medicine at Ludwig-Maximilians-Universität München (Ethikkommission bei der Medizinischen Fakultät der LMU München) under Protocol #19-872. Written informed consent was obtained from all subjects prior to the collection of human blood samples. In addition to primary samples, only commercially available established cell lines were used for this study. Funding The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This project was funded by the German Federal Ministry for Economic Affairs and Climate Action by the “Central Innovation Program for small and medium-sized enterprises (SMEs)” (ZIM) under grant numbers KK5368301AD1, KK5367301AD1, and KK5022312AD1. The TEM images of the present study were supported by the “pro patient” grant (Grant No. pp18-08) and Krebsliga 462 beider Basel (Grant No. 18-463 2016). The Orbitrap Fusion Lumos mass spectrometer was funded in part by the German Research Foundation (INST 95/1436-1 FUGG). Author Contribution DC: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft. MP: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – review & editing. GS: Funding acquisition, Project administration, Writing – review & editing. AM: Funding acquisition, Project administration, Writing – review & editing. FB: Project administration, Writing – review & editing. CL : Data curation, Software, Writing – review & editing. SW: Data curation, Software, Writing – review & editing. BK: Writing – review & editing. MY: Writing – review & editing. CZ: Writing – review & editing. LM: Funding acquisition, Project administration, Resources, Writing – review & editing. MR: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – review & editing. Acknowledgement The authors would like to thank Franziska Hackbarth for her excellent technical assistance and maintenance of mass spectrometers. References Rudd, K. E. et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study. Lancet 395 , 200–211. https://doi.org/10.1016/S0140-6736(19)32989-7 (2020). Ohnuma, T. et al. Epidemiology, Resistance Profiles, and Outcomes of Bloodstream Infections in Community-Onset Sepsis in the United States. Crit. Care Med. 51 , 1148–1158. https://doi.org/10.1097/CCM.0000000000005870 (2023). Cao, M., Wang, G. & Xie, J. Immune dysregulation in sepsis: experiences, lessons and perspectives. Cell. Death Discov . 9 , 465. https://doi.org/10.1038/s41420-023-01766-7 (2023). Kwiecinski, J. M. & Horswill, A. R. Staphylococcus aureus bloodstream infections: pathogenesis and regulatory mechanisms. Curr. Opin. Microbiol. 53 , 51–60. https://doi.org/10.1016/j.mib.2020.02.005 (2020). Urbonaviciute, V. et al. Factors masking HMGB1 in human serum and plasma. J. Leukoc. Biol. 81 , 67–74. https://doi.org/10.1189/jlb.0306196 (2007). Thavarajah, T. et al. The plasma peptides of sepsis. Clin. Proteom. 17 , 26. https://doi.org/10.1186/s12014-020-09288-5 (2020). Daigo, K. et al. The proteomic profile of circulating pentraxin 3 (PTX3) complex in sepsis demonstrates the interaction with azurocidin 1 and other components of neutrophil extracellular traps. Mol Cell Proteomics 11, M111 015073 (2012). https://doi.org/10.1074/mcp.M111.015073 Wasserman, A. et al. Septic patients presenting with apparently normal C-reactive protein: A point of caution for the ER physician. Med. (Baltim). 98 , e13989. https://doi.org/10.1097/MD.0000000000013989 (2019). Yanez-Mo, M. et al. Biological properties of extracellular vesicles and their physiological functions. J. Extracell. Vesicles . 4 , 27066. https://doi.org/10.3402/jev.v4.27066 (2015). Weber, B., Henrich, D., Hildebrand, F., Marzi, I. & Leppik, L. The Roles of Extracellular Vesicles in Sepsis and Systemic Inflammatory Response Syndrome. Shock 59, 161–172 (2023). https://doi.org/10.1097/SHK.0000000000002010 De Langhe, N. et al. Mapping bacterial extracellular vesicle research: insights, best practices and knowledge gaps. Nat. Commun. 15 , 9410. https://doi.org/10.1038/s41467-024-53279-1 (2024). You, B., Yang, Y., Wei, J., Zhou, C. & Dong, S. Pathogenic and therapeutic roles of extracellular vesicles in sepsis. Front. Immunol. 16 , 1535427. https://doi.org/10.3389/fimmu.2025.1535427 (2025). Morris, D. C. et al. Proteomic Profiles of Exosomes of Septic Patients Presenting to the Emergency Department Compared to Healthy Controls. J. Clin. Med. 9 https://doi.org/10.3390/jcm9092930 (2020). do Nascimento, M. F. et al. Circulating extracellular vesicles as potential biomarkers and mediators of acute respiratory distress syndrome in sepsis. Sci. Rep. 15 , 5512. https://doi.org/10.1038/s41598-025-89783-7 (2025). Evans, L. et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021. Crit. Care Med. 49 , e1063–e1143. https://doi.org/10.1097/CCM.0000000000005337 (2021). Benecke, L., Chiang, D. M., Ebnoether, E., Pfaffl, M. W. & Muller, L. Isolation and analysis of tumor–derived extracellular vesicles from head and neck squamous cell carcinoma plasma by galectin–based glycan recognition particles. Int. J. Oncol. 61 https://doi.org/10.3892/ijo.2022.5423 (2022). Shevchenko, A., Tomas, H., Havli, J., Olsen, J. V. & Mann, M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat. Protoc. 1 , 2856–2860. https://doi.org/10.1038/nprot.2006.468 (2006). Buchner, K., Ludwig, C., Kerpes, R. & Becker, T. Higher induction temperatures and the native secretion signal peptide promote rye prolamin 75k gamma-secalin production in Komagataella phaffii. Microb. Cell. Fact. 24 , 185. https://doi.org/10.1186/s12934-025-02809-7 (2025). Yu, M. S. C. et al. The proteome of bacterial membrane vesicles in Escherichia coli-a time course comparison study in two different media. Front. Microbiol. 15 , 1361270. https://doi.org/10.3389/fmicb.2024.1361270 (2024). Tyanova, S., Temu, T. & Cox, J. The MaxQuant computational platform for mass spectrometry-based shotgun proteomics. Nat. Protoc. 11 , 2301–2319. https://doi.org/10.1038/nprot.2016.136 (2016). Tyanova, S. et al. The Perseus computational platform for comprehensive analysis of (prote)omics data. Nat. Methods . 13 , 731–740. https://doi.org/10.1038/nmeth.3901 (2016). RStudio Integrated Development Environment for R v. 2025.05.0 + 496 (RStudio, PBC, Boston, MA). R: A language and environment for statistical computing v. 4.5.1. R Foundation for Statistical Computing, Vienna, Austria, (2025). Yu, G. Thirteen years of clusterProfiler. Innov. (Camb) . 5 , 100722. https://doi.org/10.1016/j.xinn.2024.100722 (2024). Xu, S. et al. Using clusterProfiler to characterize multiomics data. Nat. Protoc. 19 , 3292–3320. https://doi.org/10.1038/s41596-024-01020-z (2024). Wu, T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innov. (Camb) . 2 , 100141. https://doi.org/10.1016/j.xinn.2021.100141 (2021). Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16 , 284–287. https://doi.org/10.1089/omi.2011.0118 (2012). Yu, G. & He, Q. Y. ReactomePA: an R/Bioconductor package for reactome pathway analysis and visualization. Mol. Biosyst . 12 , 477–479. https://doi.org/10.1039/c5mb00663e (2016). Rohart, F., Gautier, B., Singh, A., KA, L. C. & mixOmics An R package for 'omics feature selection and multiple data integration. PLoS Comput. Biol. 13 , e1005752. https://doi.org/10.1371/journal.pcbi.1005752 (2017). MacNair, C. R. & Tan, M. W. The role of bacterial membrane vesicles in antibiotic resistance. Ann. N Y Acad. Sci. 1519 , 63–73. https://doi.org/10.1111/nyas.14932 (2023). Koike, M., Iida, K. & Matsuo, T. Electron microscopic studies on mode of action of polymyxin. J. Bacteriol. 97 , 448–452. https://doi.org/10.1128/jb.97.1.448-452.1969 (1969). Manning, A. J. & Kuehn, M. J. Contribution of bacterial outer membrane vesicles to innate bacterial defense. BMC Microbiol. 11 , 258. https://doi.org/10.1186/1471-2180-11-258 (2011). Kim, S. W. et al. Significant increase in the secretion of extracellular vesicles and antibiotics resistance from methicillin-resistant Staphylococcus aureus induced by ampicillin stress. Sci. Rep. 10 , 21066. https://doi.org/10.1038/s41598-020-78121-8 (2020). Uddin, M. J. et al. The Role of Bacterial Membrane Vesicles in the Dissemination of Antibiotic Resistance and as Promising Carriers for Therapeutic Agent Delivery. Microorganisms 8 https://doi.org/10.3390/microorganisms8050670 (2020). Andreoni, F. et al. Antibiotics Stimulate Formation of Vesicles in Staphylococcus aureus in both Phage-Dependent and -Independent Fashions and via Different Routes. Antimicrob. Agents Chemother. 63 https://doi.org/10.1128/AAC.01439-18 (2019). Kim, M. H. et al. Production of Membrane Vesicles by Enterococcus faecium Cultured With or Without Subinhibitory Concentrations of Antibiotics and Their Pathological Effects on Epithelial Cells. Front. Cell. Infect. Microbiol. 9 , 295. https://doi.org/10.3389/fcimb.2019.00295 (2019). Tian, C. et al. Extracellular vesicles participate in the pathogenesis of sepsis. Front. Cell. Infect. Microbiol. 12 , 1018692. https://doi.org/10.3389/fcimb.2022.1018692 (2022). Pepys, M. B. & Hirschfield, G. M. C-reactive protein: a critical update. J. Clin. Invest. 111 , 1805–1812. https://doi.org/10.1172/JCI18921 (2003). Xu, Y. et al. Exosomal proteome analysis of human plasma to monitor sepsis progression. Biochem. Biophys. Res. Commun. 499 , 856–861. https://doi.org/10.1016/j.bbrc.2018.04.006 (2018). Park, C. et al. Proteomic profiling of plasma extracellular vesicles identifies signatures of innate immunity, coagulation, and endothelial activation in septic patients. Sci. Rep. 15 , 21871. https://doi.org/10.1038/s41598-025-06430-x (2025). Yang, Y. X. & Li, L. Identification of potential biomarkers of sepsis using bioinformatics analysis. Exp. Ther. Med. 13 , 1689–1696. https://doi.org/10.3892/etm.2017.4178 (2017). Zetoune, F. S. & Ward, P. A. Role of Complement and Histones in Sepsis. Front. Med. (Lausanne) . 7 , 616957. https://doi.org/10.3389/fmed.2020.616957 (2020). Laarman, A., Milder, F., van Strijp, J. & Rooijakkers, S. Complement inhibition by gram-positive pathogens: molecular mechanisms and therapeutic implications. J. Mol. Med. (Berl) . 88 , 115–120. https://doi.org/10.1007/s00109-009-0572-y (2010). Broker, B. M., Mrochen, D. & Peton, V. The T Cell Response to Staphylococcus aureus. Pathogens 5 https://doi.org/10.3390/pathogens5010031 (2016). Reithmair, M. et al. Cellular and extracellular miRNAs are blood-compartment-specific diagnostic targets in sepsis. J. Cell. Mol. Med. 21 , 2403–2411. https://doi.org/10.1111/jcmm.13162 (2017). Lizak, M. & Yarovinsky, T. O. Phospholipid scramblase 1 mediates type i interferon-induced protection against staphylococcal alpha-toxin. Cell. Host Microbe . 11 , 70–80. https://doi.org/10.1016/j.chom.2011.12.004 (2012). Yan, J., Li, Z., Li, Y. & Zhang, Y. Sepsis induced cardiotoxicity by promoting cardiomyocyte cuproptosis. Biochem. Biophys. Res. Commun. 690 , 149245. https://doi.org/10.1016/j.bbrc.2023.149245 (2024). Roh, T. et al. Fucosylated haptoglobin promotes inflammation via Mincle in sepsis: an observational study. Nat. Commun. 16 , 1342. https://doi.org/10.1038/s41467-025-56524-3 (2025). Soden, M., Klett, C., Hasmann, T. & Hackenthal, E. Angiotensinogen: an acute-phase protein? Hypertension 23, I126–130 (1994). https://doi.org/10.1161/01.hyp.23.1_suppl.i126 Negron, O. et al. Fibrinogen gamma' promotes host survival during Staphylococcus aureus septicemia in mice. J. Thromb. Haemost . 21 , 2277–2290. https://doi.org/10.1016/j.jtha.2023.03.019 (2023). Iacobazzi, V. et al. Mitochondrial carriers in inflammation induced by bacterial endotoxin and cytokines. Biol. Chem. 398 , 303–317. https://doi.org/10.1515/hsz-2016-0260 (2017). Wu, W. et al. MFG-E8 has guiding significance for the prognosis and treatment of sepsis. Sci. Rep. 12 , 20916. https://doi.org/10.1038/s41598-022-25601-8 (2022). Demaret, J. et al. Identification of CD177 as the most dysregulated parameter in a microarray study of purified neutrophils from septic shock patients. Immunol. Lett. 178 , 122–130. https://doi.org/10.1016/j.imlet.2016.08.011 (2016). Timar, C. I. et al. Antibacterial effect of microvesicles released from human neutrophilic granulocytes. Blood 121 , 510–518. https://doi.org/10.1182/blood-2012-05-431114 (2013). Additional Declarations No competing interests reported. Supplementary Files SupplementalFigure1.jpg Supplemental Figure 1.(A–E) TEM images of EV complexes under different treatment conditions: mock control (A), SA (B), P100 (piperacillin–tazobactam, 100 µg/mL) (C), V25 (vancomycin, 25 µg/mL) (D), and M2.9 (moxifloxacin, 2.9 µg/mL) (E). Supplemental Figure 1. (A-E) TEM images of EV complexes under different treatment conditions. SupplementalFigure2.jpg Supplemental Figure 2. Protein counts in plasma and BacB fractions across treatment groups. Bar plots show the total number of proteins detected in plasma (A) and BacB-enriched (B) samples from the MOCK, SA, 100 µg/mL piperacillin–tazobactam (PIPTAZO), 25 µg/mL vancomycin (VAN), and 2.9 µg/mL moxifloxacin (MOXI) groups. Individual replicates are shown as points, with bars indicating the mean ± SD. QC = quality control. SupplementalFigure3.jpg Supplemental Figure 3. (A–B) Bar plots with overlaid scatter points show mean ± SEM of CMO⁺ particle concentration per mL of serum across groups (HC, BCN, BCP). Individual dots represent biological replicates (n = 3 per group). Statistical significance was assessed by one-way ANOVA. Volcano plots (C) compare protein expression between BCN vs. BCP with significantly upregulated proteins (red) and downregulated proteins (blue). Heatmaps (D) display pathway enrichment based on differentially expressed proteins in BCP vs. BCN, highlighting processes such as vesicle biogenesis, inflammation, and antimicrobial responses. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 25 Mar, 2026 Reviews received at journal 24 Mar, 2026 Reviews received at journal 18 Mar, 2026 Reviewers agreed at journal 18 Mar, 2026 Reviewers agreed at journal 17 Mar, 2026 Reviewers agreed at journal 17 Mar, 2026 Reviewers agreed at journal 01 Mar, 2026 Reviewers invited by journal 27 Feb, 2026 Editor assigned by journal 16 Jan, 2026 Submission checks completed at journal 09 Jan, 2026 First submitted to journal 09 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8384635","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":582264156,"identity":"a06e8c08-9361-4ee0-9bb3-b59cb17d3de0","order_by":0,"name":"Dapi Menglin Chiang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0ElEQVRIiWNgGAWjYDCCAxDEA2QyPoCJEK2F2eAAsVpgJJsEUVr4jp89eICh5o4M/+z2a9Uf27YBRRrwa5E8k5dwgOHYMx6JO2fKbhxsuw0UIWCNwYEcgwOMDYd5GG7kpIG1GNxIIKDl/BuIFnmglgKwlvsPCGi5AbXF4Eb6MQaILfh1MEjeeJdwIOHYYR7DGznMEmfO3eaRPEPAYXzncw9/+FBz2F7uRvrDDxVlt+X4jh8gYA0DMBIhxvIYQLkEAVwN+wMiVI+CUTAKRsFIBACJCVXcAKy5egAAAABJRU5ErkJggg==","orcid":"","institution":"Technical University of Munich (TUM)","correspondingAuthor":true,"prefix":"","firstName":"Dapi","middleName":"Menglin","lastName":"Chiang","suffix":""},{"id":582264160,"identity":"ea8cd92a-89ce-4672-a5e0-1168c98eadbb","order_by":1,"name":"Michael W. Pfaffl","email":"","orcid":"","institution":"Technical University of Munich (TUM)","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"W.","lastName":"Pfaffl","suffix":""},{"id":582264161,"identity":"1865ec32-9302-4820-926f-792c33bdddd9","order_by":2,"name":"Gustav Schelling","email":"","orcid":"","institution":"LMU University Hospital Munich","correspondingAuthor":false,"prefix":"","firstName":"Gustav","middleName":"","lastName":"Schelling","suffix":""},{"id":582264165,"identity":"325d91ba-480d-4a50-a3ec-b794bdf2dc38","order_by":3,"name":"Agnes S. Meidert","email":"","orcid":"","institution":"LMU University Hospital Munich","correspondingAuthor":false,"prefix":"","firstName":"Agnes","middleName":"S.","lastName":"Meidert","suffix":""},{"id":582264166,"identity":"492744b7-6b59-401f-bbfc-37b59a89a798","order_by":4,"name":"Florian Brandes","email":"","orcid":"","institution":"LMU University Hospital Munich","correspondingAuthor":false,"prefix":"","firstName":"Florian","middleName":"","lastName":"Brandes","suffix":""},{"id":582264175,"identity":"16bef04b-dbb8-448e-aca6-729bccc49eef","order_by":5,"name":"Christina Ludwig","email":"","orcid":"","institution":"Technical University of Munich","correspondingAuthor":false,"prefix":"","firstName":"Christina","middleName":"","lastName":"Ludwig","suffix":""},{"id":582264177,"identity":"4e215259-1319-4168-bcd9-3d9ad2335e5c","order_by":6,"name":"Susanne I. Wudy","email":"","orcid":"","institution":"Technical University of Munich","correspondingAuthor":false,"prefix":"","firstName":"Susanne","middleName":"I.","lastName":"Wudy","suffix":""},{"id":582264179,"identity":"054e5dd5-1acc-49ad-9d57-7fdc06c72050","order_by":7,"name":"Benedikt Kirchner","email":"","orcid":"","institution":"Technical University of Munich (TUM)","correspondingAuthor":false,"prefix":"","firstName":"Benedikt","middleName":"","lastName":"Kirchner","suffix":""},{"id":582264180,"identity":"299a003f-660a-443e-a92b-16b0808a7118","order_by":8,"name":"Mia S.C. Yu","email":"","orcid":"","institution":"Technical University of Munich (TUM)","correspondingAuthor":false,"prefix":"","firstName":"Mia","middleName":"S.C.","lastName":"Yu","suffix":""},{"id":582264181,"identity":"41c3c09e-7821-4533-be7d-b015b84941cf","order_by":9,"name":"Christian Zenner","email":"","orcid":"","institution":"Technical University of Munich (TUM)","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Zenner","suffix":""},{"id":582264182,"identity":"04d4d40a-0b9e-42de-ae72-1e9f93eb0989","order_by":10,"name":"Rosalie Ulbricht","email":"","orcid":"","institution":"University Hospital, LMU Munich","correspondingAuthor":false,"prefix":"","firstName":"Rosalie","middleName":"","lastName":"Ulbricht","suffix":""},{"id":582264183,"identity":"fd9f9e7c-1c0d-4123-b586-a591429a7f72","order_by":11,"name":"Laurent Muller","email":"","orcid":"","institution":"University Hospital of Basel","correspondingAuthor":false,"prefix":"","firstName":"Laurent","middleName":"","lastName":"Muller","suffix":""},{"id":582264184,"identity":"ae9ad77b-5a57-4aae-8a2d-d63983b7133f","order_by":12,"name":"Marlene Reithmair","email":"","orcid":"","institution":"University Hospital, LMU Munich","correspondingAuthor":false,"prefix":"","firstName":"Marlene","middleName":"","lastName":"Reithmair","suffix":""}],"badges":[],"createdAt":"2025-12-17 10:38:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8384635/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8384635/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103567247,"identity":"fa5e46a4-c632-4dd8-b551-29089b3a21ab","added_by":"auto","created_at":"2026-02-27 07:29:43","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":4852445,"visible":true,"origin":"","legend":"\u003cp\u003eWorkflow and analysis of plasma EVs and bMVs during \u003cem\u003eS. aureus\u003c/em\u003e (SA) infection with or without antibiotic treatment. (A) EVs and bMVs were isolated from EDTA plasma using Miltenyi and BacB beads, respectively, and analyzed by flow cytometry, transmission electron microscopy (TEM), fluorescence nanoparticle tracking analysis (F-NTA), and proteomics. (B) Representative TEM images show EV–Miltenyi bead complexes under different treatment conditions, with yellow dots indicating Miltenyi beads and blue arrows indicating vesicles (P100 = piperacillin–tazobactam 100 μg/mL, V25 = vancomycin 25 μg/mL, M2.9 = moxifloxacin 2.9 μg/mL; scale bars, 100 nm). (C–D) Bar plots with overlaid scatter points show mean ± SEM of F-NTA particle size (C) and CMO⁺ particle concentration (D) of EV–Miltenyi bead complexes across treatment groups; individual dots represent paired biological replicates (n = 3). Statistical significance was assessed using paired Wilcoxon matched-pairs signed-rank tests for all pairwise comparisons; piperacillin–tazobactam is abbreviated as Pip-Tazo. (E–F) Bar plots with overlaid scatter points show mean ± SEM percentages of (E) CMD⁺ PanEV⁺ CD45⁺ anti-SA⁺ EV–Miltenyi bead complexes and (F) CMD⁺ PanEV⁻ anti-SA⁺ bMV–BacB bead complexes quantified by flow cytometry across treatment groups; individual dots represent paired biological replicates (n = 6). Statistical significance was assessed using paired Wilcoxon matched-pairs signed-rank tests.\u003c/p\u003e","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/b14e787a2829b79230107dab.jpg"},{"id":103567252,"identity":"731abd12-ed5b-4fa5-ba07-dfef83c60926","added_by":"auto","created_at":"2026-02-27 07:29:43","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2247113,"visible":true,"origin":"","legend":"\u003cp\u003eProteomic profiling of plasma EVs reveals SA infection- and antibiotic-specific signatures. (A) sPLS-DA plot (components 1 vs. 2) showing distinct clustering of EV proteomes from MOCK, SA-infected, and antibiotic-treated groups (n=6 per group). Ellipses represent 95% confidence intervals. (B) Top \u0026nbsp;20 proteins contributing to component 1 separation, with highest loading values from SA-infected samples. (C) Volcano plot comparing SA vs. MOCK groups, highlighting significantly upregulated (red) and downregulated (blue) proteins. (D) Venn diagram showing overlap between top 20 proteins from component 1 and significantly changed proteins in SA vs. MOCK comparison. (E) Pathway enrichment analysis of shared significant EV proteins from sepsis samples, showing enrichment in secretion, vesicle trafficking, and immune response pathways. (F–H) Volcano plots comparing EV proteomes of SA vs. piperacillin-tazobactam (F), SA vs. vancomycin (G), and SA vs. Moxifloxacin (H) groups, showing antibiotic-specific proteomic shifts.\u003c/p\u003e","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/a613098205c3af2c1a1e3f5b.jpg"},{"id":103567250,"identity":"1e8e1162-d1dc-4522-888b-bda531797885","added_by":"auto","created_at":"2026-02-27 07:29:43","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":6539927,"visible":true,"origin":"","legend":"\u003cp\u003eSerum EV proteomes distinguish HC, BCN, and BCP groups. sPLS-DA (A) shows distinct clustering of EV proteomes from HC, BCN, and BCP samples (\u003cem\u003en\u003c/em\u003e= 6 per group). Volcano plots (B, D, G) compare protein expression between BCN vs. HC and BCP vs. HC, with significantly upregulated proteins in red and downregulated proteins in blue. Heatmaps (C, E, H) display pathway enrichment based on differentially expressed proteins in BCN vs. HC, BCP vs. HC and Bac_infection_group vs. HC, highlighting processes such as vesicle biogenesis, inflammation, and antimicrobial responses. A Venn diagram (F) illustrates the overlap of significant proteins among HC, BCN, and BCP.\u003c/p\u003e","description":"","filename":"Figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/3fc063135873cb506963c60a.jpg"},{"id":104407534,"identity":"22cb2876-428e-4cf6-b463-7874b13271cd","added_by":"auto","created_at":"2026-03-11 12:38:35","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":14477292,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/f9d08084-dfc9-4500-af4d-9f18f8077f4e.pdf"},{"id":104398563,"identity":"17ab2299-e616-42a3-a06e-4b5bc0dced48","added_by":"auto","created_at":"2026-03-11 12:02:56","extension":"jpg","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1437299,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplemental Figure 1.\u003c/strong\u003e(A–E) TEM images of EV complexes under different treatment conditions: mock control (A), SA (B), P100 (piperacillin–tazobactam, 100 µg/mL) (C), V25 (vancomycin, 25 µg/mL) (D), and M2.9 (moxifloxacin, 2.9 µg/mL) (E). Supplemental Figure 1. (A-E) TEM images of EV complexes under different treatment conditions.\u003c/p\u003e","description":"","filename":"SupplementalFigure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/7bcbe7d09196322fd6801497.jpg"},{"id":104398095,"identity":"6668cadf-11d1-453a-809b-670c847ee5ef","added_by":"auto","created_at":"2026-03-11 11:59:43","extension":"jpg","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":577380,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplemental Figure 2.\u003c/strong\u003e Protein counts in plasma and BacB fractions across treatment groups. Bar plots show the total number of proteins detected in plasma (A) and BacB-enriched (B) samples from the MOCK, SA, 100 µg/mL piperacillin–tazobactam (PIPTAZO), 25 µg/mL vancomycin (VAN), and 2.9 µg/mL moxifloxacin (MOXI) groups. Individual replicates are shown as points, with bars indicating the mean ± SD. QC = quality control.\u003c/p\u003e","description":"","filename":"SupplementalFigure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/250a27f501619b34756c63e1.jpg"},{"id":104397898,"identity":"12c7b1d4-1aff-4fc8-b8e7-f41f2f0eb7ed","added_by":"auto","created_at":"2026-03-11 11:58:45","extension":"jpg","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":3486228,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplemental Figure 3.\u003c/strong\u003e (A–B) Bar plots with overlaid scatter points show mean ± SEM of CMO⁺ particle concentration per mL of serum across groups (HC, BCN, BCP). Individual dots represent biological replicates (n = 3 per group). Statistical significance was assessed by one-way ANOVA. Volcano plots (C) compare protein expression between BCN vs. BCP with significantly upregulated proteins (red) and downregulated proteins (blue). Heatmaps (D) display pathway enrichment based on differentially expressed proteins in BCP vs. BCN, highlighting processes such as vesicle biogenesis, inflammation, and antimicrobial responses.\u003c/p\u003e","description":"","filename":"SupplementalFigure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8384635/v1/3337335309262274adc389a0.jpg"}],"financialInterests":"No competing interests reported.","formattedTitle":"Host-Defense Extracellular Vesicle Protein Changes in Antibiotic- and Staphylococcus aureus–Treated Blood","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSepsis is a severe, life-threatening syndrome in which the host\u0026rsquo;s response to infection becomes dysregulated, triggering systemic inflammation, tissue injury, and the risk of multiple organ failure\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Sepsis continues to affect millions worldwide, with about 49\u0026nbsp;million cases and 11\u0026nbsp;million deaths each year, and the burden falls most heavily on low- and middle-income nations\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Sepsis commonly develops when bacteria from a localized infection enter the bloodstream. In community-onset sepsis, frequent bloodstream pathogens include \u003cem\u003eEscherichia coli\u003c/em\u003e, \u003cem\u003eKlebsiella\u003c/em\u003e species, \u003cem\u003eStreptococcus\u003c/em\u003e species, and \u003cem\u003eEnterococcus\u003c/em\u003e species\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eStaphylococcus aureus\u003c/em\u003e (\u003cem\u003eSA\u003c/em\u003e), including both methicillin-susceptible (MS) SA and methicillin-resistant (MR) SA strains, respectively MSSA and MRSA, is often found in blood cultures, and MRSA infections are associated with higher hospital mortality\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. In sepsis involving SA, immune homeostasis becomes severely disrupted, with strong inflammatory activation occurring alongside an immunosuppressive state that allows the bacteria to persist and spread\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn early sepsis, some well known protein biomarker like S100A8/A9, pentraxin-3 (PTX3), serum amyloid A1 (SAA1), high mobility group box 1 (HMGB1), resistin, and complement C3/C5 are elevated and reflect disease severity. These key host responses including neutrophil, endothelial, and complement activation, although proteins like SAA1 and PTX3 are difficult to detect because of low baseline levels, rapid fluctuations during illness, and interference from abundant plasma proteins\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. In contrast, clinically used markers such as C-reactive protein (CRP) are nonspecific and can give a false sense of reassurance, since CRP may remain low even in confirmed sepsis, particularly in the early phase of illness or in immunocompromised patients, which can delay timely diagnosis and treatment\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Together, these issues reveal the constraints of relying solely on untargeted plasma proteomics and suggest that extracellular vesicles (EVs) may provide a richer and more mechanistically relevant pool of biomarkers.\u003c/p\u003e \u003cp\u003eEVs released by human cells, particularly small EVs (sEVs, 30\u0026ndash;150 nm) formerly known as exosomes, carry cargo that reflects their cellular origin, and unlike low-abundance or rapidly cleared plasma proteins, they provide a concentrated and protected source of mechanistically informative mediators, making them valuable players and biomarker candidates in sepsis\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Bacteria also secrete vesicles of similar size, known as bacterial membrane vesicles (bMVs), which carry virulence factors and immunomodulatory molecules that can contribute to sepsis pathogenesis\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. In sepsis, sEVs are actively released by immune cells, endothelial cells, epithelial cells, and platelets in response to infection, hypoxia, and systemic inflammation, contributing to both the propagation and regulation of host immune responses\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Proteome profiles of sEVs in sepsis patients often show higher levels of inflammation-related proteins, including SAA1, CRP, and immunoglobulin components, suggesting that these vesicles reflect key inflammatory processes and may aid in diagnosis\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Monocyte- and platelet-derived EVs in sepsis carry surface markers and cargo that promote endothelial activation, increase vascular permeability, and enhance tissue factor\u0026ndash;mediated coagulation, directly contributing to microvascular dysfunction and multi-organ failure\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Although interest in sEVs as biomarkers in sepsis is growing, the dynamic response of the host sEVs proteome to direct bacterial challenges, such as \u003cem\u003eSA\u003c/em\u003e infection, remains poorly understood. Furthermore, how antibiotic treatment shapes this response remains unclear because comprehensive comparisons among culture-positive sepsis, culture-negative sepsis, and healthy individuals are limited, even though such analyses are essential for defining diagnostic specificity and clinical usefulness. In this study, we investigated the human EV proteomic landscape using a dual approach: (1) an \u003cem\u003eex vivo\u003c/em\u003e whole-blood infection model challenged with \u003cem\u003eSA\u003c/em\u003e, with or without antibiotic treatment, and (2) clinical plasma samples from patients with blood culture\u0026ndash;positive and \u0026ndash;negative sepsis, alongside healthy volunteers. Using immunomagnetic bead\u0026ndash;based EV isolation together with label-free proteomics and multivariate analysis, we set out to define EV signatures that track with infection status and treatment effects. This approach allowed us to detect both human and bacterial proteins within the vesicles and to observe how antibiotic exposure influenced their composition. Building on these findings, we compared EV proteomes from culture-positive sepsis, culture-negative sepsis, and healthy individuals to define EV signatures that reflect infection status and treatment response in sepsis.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSA-Spiked Plasma Sample Collection\u003c/h2\u003e \u003cp\u003ePeripheral blood (55 ml) was collected from six healthy volunteers in EDTA-coated tubes to prevent coagulation. From each donor, 5 ml of whole blood was distributed into 15 ml Falcon tubes and allocated to 11 experimental conditions: a MOCK (non-infection) control, infection with \u003cem\u003eSA\u003c/em\u003e DSM 20231 (MSSA), and infection in combination with nine antibiotic treatments. The \u003cem\u003eSA\u003c/em\u003e DSM 20231 strain was obtained from DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen, German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany) and maintained on tryptic soy broth (TSB) agar plates. A single colony was inoculated into 500 ml of RPMI-1640 medium, prepared as two 250-ml cultures, and grown at 37\u0026deg;C, yielding a bacterial suspension with an OD₆₀₀ of approximately 1.0\u0026ndash;1.5. From the resulting culture mixture, a 1-ml aliquot was labeled with the CFSE Cell Division Tracker Kit (BioLegend, San Diego, USA) for 30 min at 37\u0026deg;C. CFSE-labeled \u003cem\u003eSA\u003c/em\u003e were quantified by flow cytometry (LSRFortessa\u0026trade;, Becton, Dickinson and Company, Franklin Lakes, NJ, USA), and data analysis was performed using FlowJo 10.8 (Becton, Dickinson and Company, Franklin Lakes, NJ, USA).\u003c/p\u003e \u003cp\u003eFor infection, 3 \u0026times; 10⁷ \u003cem\u003eSA\u003c/em\u003e were added to 5 ml of whole blood, corresponding to a multiplicity of infection (MOI) of 0.001 and approximately 6 \u0026times; 10⁹ cells/ml as measured with a TC10\u0026trade; Automated Cell Counter (Bio-Rad). Antibiotic concentrations were selected to represent clinically relevant plasma levels: piperacillin\u0026ndash;tazobactam (25, 50, 100 \u0026micro;g/ml), vancomycin (6.25, 12.5, 25 \u0026micro;g/ml), and moxifloxacin (0.18125, 0.725, 2.9 \u0026micro;g/ml). These antibiotics were selected for their distinct mechanisms of action, including inhibition of cell wall synthesis by piperacillin\u0026ndash;tazobactam (Pip-Tazo) and vancomycin, and inhibition of DNA replication by moxifloxacin, in accordance with the Surviving Sepsis Campaign 2021 international guidelines for empiric treatment of \u003cem\u003eSA\u003c/em\u003e and sepsis\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. All 11 experimental conditions were used for flow cytometric analysis, whereas only the highest concentration of each of the three antibiotics (100 \u0026micro;g/ml Pip-Tazo, 25 \u0026micro;g/ml vancomycin, and 2.9 \u0026micro;g/ml moxifloxacin), together with the MOCK control and infection-alone condition, was used for subsequent proteomic analyses. Following 24 h incubation at 37\u0026deg;C, plasma was separated by centrifugation at 2,000 \u0026times; g for 10 min at 4\u0026deg;C and filtered through a 0.45 \u0026micro;m PES syringe filter (Merck KGaA, Darmstadt, Germany) to remove residual bacteria. Approximately 2 ml of filtered plasma was collected per condition and stored at \u0026minus;\u0026thinsp;80\u0026deg;C until analysis.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eHuman EV isolation\u003c/h3\u003e\n\u003cp\u003eHuman EVs were isolated from (i) 1 ml of filtered plasma obtained from the SA-spiked blood model (MOCK control, infection-only, and antibiotic-treated conditions; n\u0026thinsp;=\u0026thinsp;6 per condition) and (ii) serum EV\u0026ndash;Miltenyi bead complexes collected from healthy controls (HC, n\u0026thinsp;=\u0026thinsp;6), bacterial culture\u0026ndash;negative patients (BCN, n\u0026thinsp;=\u0026thinsp;6), and bacterial culture\u0026ndash;positive patients (BCP, n\u0026thinsp;=\u0026thinsp;6). EV isolation was performed using immunoaffinity capture with the EV Isolation Kit Pan, human (Miltenyi Biotec, Bergisch Gladbach, Germany), following the manufacturer\u0026rsquo;s protocol. Briefly, 1 ml of filtered plasma or serum was incubated with 50 \u0026micro;l of immunoaffinity beads for 1 hour on an overhead rotor. A Miltenyi \u0026micro;Column was equilibrated and washed three times before sample loading, followed by four additional washes on a magnetic stand. EV\u0026ndash;Miltenyi bead complexes were eluted using 100 \u0026micro;l of the kit-supplied isolation buffer. The unbound plasma fraction (1 ml) collected during the immunoaffinity procedure was retained for additional downstream analyses. Clinical characteristics of all participants, including age, sex, hospital and intensive care stay, pathogen distribution, day 1 antibiotic treatment status, and day 1 inflammatory markers (CRP, interleukin-6, leukocyte count, and procalcitonin), are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of clinical characteristics.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBCP (n\u0026thinsp;=\u0026thinsp;6)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBCN (n\u0026thinsp;=\u0026thinsp;6)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (years) \u0026ndash; Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e67.8\u0026thinsp;\u0026plusmn;\u0026thinsp;16.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e74.8\u0026thinsp;\u0026plusmn;\u0026thinsp;15.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHospital stay \u0026ndash; Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.8\u0026thinsp;\u0026plusmn;\u0026thinsp;2.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9.3\u0026thinsp;\u0026plusmn;\u0026thinsp;6.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIntensive care \u0026ndash; Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.0\u0026thinsp;\u0026plusmn;\u0026thinsp;2.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.7\u0026thinsp;\u0026plusmn;\u0026thinsp;3.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePathogens\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eSA\u003c/em\u003e (4), \u003cem\u003eS. capitis\u003c/em\u003e (1), \u003cem\u003eC. tertium\u003c/em\u003e (1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNone\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay 1 antibiotic treatment \u0026ndash; Yes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay 1 antibiotic treatment \u0026ndash; No\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay 1 C-reactive protein\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22.7\u0026thinsp;\u0026plusmn;\u0026thinsp;16.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22.4\u0026thinsp;\u0026plusmn;\u0026thinsp;14.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay 1 Interleukin-6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e567.9\u0026thinsp;\u0026plusmn;\u0026thinsp;590.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e199.8\u0026thinsp;\u0026plusmn;\u0026thinsp;203.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay 1 Leukocytes count\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11.1\u0026thinsp;\u0026plusmn;\u0026thinsp;7.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20.2\u0026thinsp;\u0026plusmn;\u0026thinsp;13.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay 1 Procalcitonin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13.9\u0026thinsp;\u0026plusmn;\u0026thinsp;12.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.9\u0026thinsp;\u0026plusmn;\u0026thinsp;3.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch3\u003eIn-house bMV isolation\u003c/h3\u003e\n\u003cp\u003eThe bacterial beads (BacB) necessary for the bMVs isolation is described in brief: First, 250 \u0026micro;l of 40 \u0026micro;g/ml 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Thermo Fisher Scientific, Bremen, Germany) in 0.1 M 2-(N-morpholino)ethanesulfonic acid (MES) buffer (Sigma\u0026thinsp;+\u0026thinsp;Merck KGaA, Darmstadt, Germany), pH 5.0, was used to activate 10 mg of carboxyl-group magnetic particles (2.13 \u0026times; 10\u0026sup1;⁰ particles/ml, IKERLAT Polymers, Lasarte-Oria, Spain) at 25\u0026deg;C for 15 minutes. The activated magnetic particles were then resuspended in 1 ml of PBS. Next, 50 \u0026micro;g each of OmpA antibodies (catalog number: orb26813, Biorbyt, Cambridge, UK) and GroEL antibodies (catalog number: NBP2-32867, Novus Biologicals, Centennial, United States) were prepared. A total of 100 \u0026micro;l of the antibody mixture was incubated with 1 ml of activated magnetic particles at 37\u0026deg;C for 2 hours. To block unoccupied binding sites, the antibody-coated beads were incubated with 1 ml of 10% EV-free BSA in PBS (prepared by centrifugation at 100,000 \u0026times; g, 4\u0026deg;C, overnight) at 37\u0026deg;C for 30 minutes. The resulting beads were then diluted to 1 \u0026times; 10⁶ beads/\u0026micro;l and used as BacB. One milliliter of unbound plasma was incubated with 100 \u0026micro;l of BacB at 37\u0026deg;C for 1 hour. After incubation, bMV\u0026ndash;BacB complexes were washed twice with 1 ml of 0.5% EV-free BSA in PBS while magnetic bMV\u0026ndash;BacB complexes were retained in the tube using a magnetic stand. As a negative control, 1 ml of 10% EV-free (EF) BSA in PBS was incubated with BacB under the same conditions to generate EF\u0026ndash;BacB complexes. bMV\u0026ndash;BacB complexes were used for further analysis by bead-based flow cytometry or proteomics, while EF\u0026ndash;BacB complexes were used only as a negative control for bead-based flow cytometry.\u003c/p\u003e\n\u003ch3\u003eBead-based flow cytometry\u003c/h3\u003e\n\u003cp\u003e25 \u0026micro;l of EV\u0026ndash;Miltenyi bead complexes or 250 \u0026micro;l of bMV\u0026ndash;BacB complexes were diluted in 0.9 ml of 0.5% EV-free BSA in PBS. To this mixture, 100 \u0026micro;l of an antibody master mix was added. The master mix contained 2.5 \u0026micro;l of PanEV antibody (3 \u0026micro;g), pre-mixed with 1 \u0026micro;l each of PE-conjugated anti-CD63 (clone H5C6), PE-conjugated anti-CD81 (clone 5A6), and PE-conjugated anti-CD9 (clone HI9a) (all from BioLegend, San Diego, USA); 5 \u0026micro;l of Brilliant Violet 421\u0026trade; anti-human CD45 antibody (clone HI30, BioLegend, San Diego, USA); 10 \u0026micro;M CellMask\u0026trade; Plasma Membrane Deep Red (CMD; C10046, Thermo Fisher Scientific, Bremen, Germany); and 1 \u0026micro;l of FITC-conjugated \u003cem\u003eSA\u003c/em\u003e polyclonal antibody (PA1-73172, Thermo Fisher Scientific, Bremen, Germany). The mixture was incubated at 37\u0026deg;C for 1 hour. After staining, EV\u0026ndash;Miltenyi bead complexes and bMV\u0026ndash;BacB complexes were washed twice with 1 ml of 0.5% EV-free BSA in PBS and centrifuged at 100,000 \u0026times; g at 4\u0026deg;C overnight using a magnetic rack. The samples were then analyzed on a BD LSR Fortessa\u0026trade; flow cytometer (Becton, Dickinson and Company, Franklin Lakes, USA). Data analysis, including gating, was performed using FlowJo software version 10.8 (Becton, Dickinson and Company, Franklin Lakes, USA). An IgG isotype control was included to confirm the specificity of the staining. In the flow cytometry gating strategy, EV\u0026ndash;Miltenyi bead complexes and bMV\u0026ndash;BacB complexes were initially gated based on forward scatter (FSC-A) and side scatter (SSC-A). The CMD⁺ PanEV⁺ CD45⁺ Anti-SA⁺ population from the EV\u0026ndash;Miltenyi bead complexes was identified as immune cell-derived EVs carrying \u003cem\u003eSA\u003c/em\u003e antigen. In contrast, the CMD⁺ PanEV⁻ Anti-SA⁺ population from the bMV\u0026ndash;BacB complexes was identified based on comparison to the isotype control.\u003c/p\u003e\n\u003ch3\u003eHuman EV characterization\u003c/h3\u003e\n\u003cp\u003eNanoparticle tracking analysis (NTA) was used to determine the size distribution and particle concentration of extracellular vesicles (EVs) using ZetaView\u0026reg; PMX 110 or x30 MONO instruments equipped with a 520-nm laser and a 550-nm emission filter (Particle Metrix GmbH, Inning am Ammersee, Germany). EV samples were diluted in particle-free PBS to fall within the optimal measurement range of the instrument and were measured in two acquisition cycles across 11 positions per sample. For fluorescence measurements, EV membranes were labeled with CellMask Orange\u0026trade; (CMO; 5 mg/mL stock; Thermo Fisher Scientific Inc., Waltham, MA, USA) at a final concentration of 25 \u0026micro;g/mL (1:200 dilution), and samples were further diluted 1:400 prior to analysis unless otherwise stated. Fluorescence measurements on the PMX 110 system were performed using 1 mL of filtered plasma obtained from a \u003cem\u003eStaphylococcus aureus\u003c/em\u003e\u0026ndash;spiked blood model (mock control, infection-only, and antibiotic-treated conditions; n\u0026thinsp;=\u0026thinsp;6 per group) with video capture settings of shutter speed 70, sensitivity 95, and a frame rate of 30 frames/s, and data were analyzed using ZetaView software version 8.05.11 SP1. Fluorescence measurements on the x30 MONO system were conducted on serum EV\u0026ndash;Miltenyi bead complexes from healthy controls (HC; n\u0026thinsp;=\u0026thinsp;6), bacterial culture-negative patients (BCN; n\u0026thinsp;=\u0026thinsp;6), and bacterial culture-positive patients (BCP; n\u0026thinsp;=\u0026thinsp;6) using a shutter speed of 100, sensitivity 90, and a frame rate of 30 frames/s, with data analyzed using ZetaView software version 8.06.01 SP1 (Particle Metrix GmbH, Inning am Ammersee, Germany).\u003c/p\u003e \u003cp\u003eCryo-electron microscopy (cryo-EM) sample preparation was conducted as previously described\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. In brief, 4 \u0026micro;l of each sample was applied to glow-discharged holey carbon-coated grids (Ted Pella Inc., Redding, United States), blotted using Whatman No. 1 filter paper, and vitrified by plunge-freezing into liquid ethane at \u0026minus;\u0026thinsp;178\u0026deg;C with a Leica GP plunger (Leica Microsystems GmbH, Wetzlar, Germany). Vitrified grids were transferred to a Talos electron microscope (FEI/Thermo Fisher Scientific Inc., Waltham, United States) using a Gatan 626 cryo-holder (Gatan Inc., Pleasanton, United States). Imaging was performed at an accelerating voltage of 200 kV under low-dose conditions (20 e⁻/\u0026Aring;\u0026sup2;) while maintaining cryogenic temperatures. Micrographs were acquired using a CETA camera (Thermo Fisher Scientific Inc., Waltham, United States).\u003c/p\u003e \u003cp\u003eNegative staining transmission electron microscopy (TEM) was performed as previously described\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Briefly, 5 \u0026micro;l of undiluted sample was adsorbed for 60 seconds onto glow-discharged parlodion/carbon-coated copper grids (FCF200-Ni; Electron Microscopy Sciences, Hatfield, United States). The grids were then blotted, rinsed three times with double-distilled water (ddH₂O), and stained on two consecutive droplets of 2% uranyl acetate solution (Merck KGaA, Darmstadt, Germany). Imaging was carried out using a Talos F200C TEM (FEI/Thermo Fisher Scientific Inc., Waltham, United States) operated at an accelerating voltage of 120 kV. Electron micrographs were acquired using a Veleta camera (EMSIS GmbH, M\u0026uuml;nster, Germany).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eProtein preparation for proteomics\u003c/h2\u003e \u003cp\u003eEV proteins from (i) plasma EV\u0026ndash;Miltenyi bead complexes, unbound plasma bMV\u0026ndash;BacB complexes, and total plasma obtained from the SA-spiked blood model (MOCK control, infection-only, and antibiotic-treated conditions; n\u0026thinsp;=\u0026thinsp;6 per condition), and from (ii) serum EV\u0026ndash;Miltenyi bead complexes collected from healthy controls (HC, n\u0026thinsp;=\u0026thinsp;6), bacterial culture\u0026ndash;negative (BCN, n\u0026thinsp;=\u0026thinsp;6), and bacterial culture\u0026ndash;positive (BCP, n\u0026thinsp;=\u0026thinsp;6) patient samples were lysed in 1\u0026times; RIPA buffer (Abcam plc, Cambridge, United Kingdom; ab156034) supplemented with 1\u0026times; ProteaseArrest\u0026trade; Protease Inhibitor Cocktail (G-BIOSCIENCES, St. Louis, MO, United States), and stored at \u0026minus;\u0026thinsp;80\u0026deg;C until further use. Protein samples were then boiled at 70\u0026deg;C for 10 minutes and sonicated on ice for 5 minutes. Following sonication, lysates were centrifuged at 10,000 \u0026times; g at 4\u0026deg;C for 30 minutes. Protein concentrations were determined using the bicinchoninic acid (BCA) assay (Thermo Fisher Scientific Inc., Waltham, United States). A minimum of 5 \u0026micro;g of protein from each sample was mixed with 1\u0026times; Laemmli Sample Buffer (Bio-Rad Laboratories, Hercules, United States) containing 2-mercaptoethanol (Merck KGaA, Darmstadt, Germany) in a final volume of 35 \u0026micro;l. Prepared samples were then submitted to the proteomics core facility at the Bavarian Center for Biomolecular Mass Spectrometry for analysis.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMass spectrometry-based proteomics\u003c/h3\u003e\n\u003cp\u003eIn accordance with standard procedures, in-gel trypsin digestion was performed\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e on all samples derived from both experimental datasets. This included (i) EV\u0026ndash;Miltenyi bead complexes, bMV\u0026ndash;BacB complexes, and total plasma obtained from the SA-spiked blood model (MOCK control, infection-only, and antibiotic-treated conditions; n\u0026thinsp;=\u0026thinsp;6 per condition), as well as (ii) EVs and total plasma or serum collected from healthy controls (HC, n\u0026thinsp;=\u0026thinsp;6), bacterial culture\u0026ndash;negative (BCN, n\u0026thinsp;=\u0026thinsp;6), and bacterial culture\u0026ndash;positive (BCP, n\u0026thinsp;=\u0026thinsp;6) patient samples. Briefly, each sample was loaded onto a NuPAGE\u0026trade; 4\u0026ndash;12% Bis-Tris protein gel (Thermo Fisher Scientific, Waltham, USA) and run approximately 1 cm into the gel to concentrate the proteins into a single non\u0026ndash;size-separated band. The accumulated band was excised, reduced with 50 mM dithiothreitol, alkylated with 55 mM chloroacetamide, and digested overnight at 37\u0026deg;C with Trypsin Gold (mass spectrometry grade, Promega). Following digestion, peptides were extracted, dried, and resuspended in 25 \u0026micro;l of buffer A (2% acetonitrile, 0.1% formic acid in HPLC-grade water). A 5 \u0026micro;l aliquot of each peptide solution was injected for LC\u0026ndash;MS/MS analysis.\u003c/p\u003e \u003cp\u003eLC-MS/MS measurements were conducted using a Dionex Ultimate 3000 RSLCnano system linked to an Orbitrap Fusion LUMOS instrument (ThermoFisher Scientific, Bremen), in adherence to the standard protocol of the core facility\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Samples were initially loaded onto a self-packed trap column (ReproSil-pur C18-AQ, 5 \u0026micro;m, 20 mm \u0026times; 75 \u0026micro;m, Dr. Maisch) at a flow rate of 5 \u0026micro;L/min using 0.1% formic acid. After a 10-minute loading step, peptides were eluted onto a self-packed analytical column (ReproSil Gold C18-AQ, 3 \u0026micro;m, 450 mm \u0026times; 75 \u0026micro;m, Dr. Maisch). Chromatographic separation occurred over a 50-minute gradient increasing from 4% to 32% solvent B (acetonitrile containing 0.1% FA and 5% DMSO) against solvent A (water containing 0.1% FA and 5% DMSO) at a flow rate of 300 nL/min. The mass spectrometer operated in positive mode using data-dependent acquisition (DDA). We acquired full MS1 scans (360\u0026ndash;1300 m/z) at a resolution of 60,000, with a normalized AGC target of 100% and a maximum injection time of 50 ms. A 2-second cycle time was employed, selecting precursors with charge states between 2 and 6, while applying a 30-second dynamic exclusion. Fragmentation was carried out via higher-energy collision-induced dissociation (HCD) at 30% normalized collision energy (NCE) with a 1.3 m/z isolation window. MS2 spectra were recorded at a resolution of 15,000, using an AGC target of 150% and a maximum injection time of 22 ms.\u003c/p\u003e\n\u003ch3\u003eLabel-Free Quantitative Proteomics and Integrated Statistical Analysis for Host–Pathogen Profiling in Sepsis\u003c/h3\u003e\n\u003cp\u003eProteomic peptide identification and quantification were executed in accordance with our previous study\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, utilizing the MaxQuant computational platform (version 1.6.3.4) and the integrated Andromeda search algorithm\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Tandem mass spectrometry (MS/MS) spectra were queried against UniProt reference proteomes for \u003cem\u003eHomo sapiens\u003c/em\u003e (Taxon ID: 9606, UP000005640, July 2020) and \u003cem\u003eStaphylococcus aureus\u003c/em\u003e (Taxon ID: 1280, July 2024), with the search space augmented by the standard MaxQuant contaminant database. Search parameters were configured with Trypsin/P specificity; carbamidomethylation of cysteine was defined as a fixed modification, while methionine oxidation and N-terminal acetylation were treated as variable modifications. To ensure statistical reliability, a target-decoy strategy utilizing reversed sequences was applied, restricting the false discovery rate (FDR) to 1% at both the peptide spectrum match (PSM) and protein levels. Subsequent data curation was conducted in Perseus (version v2.0.10.0)\u003csup\u003e21\u003c/sup\u003e. The dataset was filtered to exclude proteins identified merely by site, reverse hits, and potential contaminants, after which label-free quantification (LFQ) intensities underwent log10 transformation\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo identify proteins reproducibly detected across sample types including EV\u0026ndash;Miltenyi bead complexes, bMV\u0026ndash;BacB complexes, total plasma, and patient serum, proteins were retained if they had valid values in at least 50% of biological replicates within any one group, using the \u0026ldquo;Filter rows based on valid values\u0026rdquo; function in Perseus\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Missing values were imputed using the standard normal distribution\u0026ndash;based imputation in Perseus, followed by quantile normalization using the built-in Perseus normalization function. These preprocessing steps were applied consistently to both datasets: (i) the \u003cem\u003eSA\u003c/em\u003e\u0026ndash;spiked plasma experiment and (ii) the patient cohort comprising HC, BCN, and BCP groups.\u003c/p\u003e \u003cp\u003ePreprocessed data were subsequently analyzed in RStudio (version: 2025.05.0\u0026thinsp;+\u0026thinsp;496)\u003csup\u003e22\u003c/sup\u003e with R (version: 4.5.1)\u003csup\u003e23\u003c/sup\u003e. Differential expression analysis was performed separately for the two datasets. For the \u003cem\u003eSA\u003c/em\u003e\u0026ndash;spiked plasma experiment (i), paired two-sided t-tests were used to account for matched biological replicates from the same donors across conditions. For the patient cohort (ii), unpaired two-sided Student\u0026rsquo;s t-tests were applied, reflecting the independence of clinical samples. In both analyses, proteins were considered significantly regulated if they exhibited p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and an absolute log₁₀ fold change (|log₁₀FC|) greater than 0.301, corresponding to a fold change greater than 2 or less than 0.5.\u003c/p\u003e \u003cp\u003eVolcano plots were generated using ggplot2, and significant proteins were annotated using a combined human and \u003cem\u003eSA\u003c/em\u003e protein database. Protein-level functional enrichment analysis was performed using the clusterProfiler (version: 4.14.6)\u003csup\u003e24\u0026ndash;27\u003c/sup\u003e, ReactomePA (version: 1.50.0)\u003csup\u003e28\u003c/sup\u003e, and org.Hs.eg.db (version: 3.20.0) packages to identify enriched Gene Ontology (GO) terms, KEGG pathways, and Reactome pathways. Enrichment significance was evaluated using Fisher\u0026rsquo;s exact test. Results were visualized via pathway\u0026ndash;protein heatmaps, filtered using sepsis-related biological keywords. To assess global expression trends and group separation, sparse Partial Least Squares Discriminant Analysis (sPLS-DA) was performed using the mixOmics package (version: 6.30.0)\u003csup\u003e29\u003c/sup\u003e. Component-based plots with group-specific clustering and confidence ellipses were generated in RStudio (version 2025.05.0\u0026thinsp;+\u0026thinsp;496)\u003csup\u003e22\u003c/sup\u003e with R (version 4.5.1)\u003csup\u003e23\u003c/sup\u003e. Finally, Venn diagrams were constructed using the Venn Diagram package to visualize the overlap and uniqueness of differentially expressed proteins across clinical conditions.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTo assess the impact of SA infection, we isolated EVs and bMVs from plasma using two approaches: Miltenyi beads for human EV capture and BacB-coated magnetic beads for bMV\u0026ndash;EV complex enrichment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Fluorescence nanoparticle tracking analysis (F-NTA) and negative-stain electron microscopy of the EV\u0026ndash;Miltenyi bead complexes first showed no observable differences across all experimental groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB\u0026ndash;D). Negative-stain electron microscopy of bMV-BacB complexes also showed no observable differences across all the groups (Supplemental Fig.\u0026nbsp;1A-E). Analysis of the EV\u0026ndash;Miltenyi bead complexes revealed a significant increase in CMD⁺ PanEV⁺ CD45⁺ \u003cem\u003eSA\u003c/em\u003e\u003csup\u003e\u003cem\u003e+\u003c/em\u003e\u003c/sup\u003e signals in the SA-spiked group compared to the MOCK control (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.032) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). Within the \u003cem\u003eSA\u003c/em\u003e-spiked groups, treatment with either the lowest concentration of piperacillin-tazobactam (25 \u0026micro;g/ml) (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.032) or the highest concentration of vancomycin (25 \u0026micro;g/ml) (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.032) also resulted in significantly higher CMD⁺ PanEV⁺ CD45⁺ SA⁺ signals compared to the SA-spiked group without antibiotics (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). Analysis of the bMV fraction, using unbound plasma collected after Miltenyi bead isolation, revealed that the bMV\u0026ndash;BacB complexes exhibited a significant increase in CMD⁺ PanEV⁺ SA⁺ signals in the \u003cem\u003eSA\u003c/em\u003e-spiked group compared to the MOCK control (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.032) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF). However, no significant differences were observed between the antibiotic-treated groups and the \u003cem\u003eSA\u003c/em\u003e-spiked group without treatment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF). These findings suggest that \u003cem\u003eSA\u003c/em\u003e infection induces significant changes in the protein composition of EVs and bMVs, rather than altering their total quantity, as indicated by EV flow cytometry and F-NTA.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo better understand how SA infection with or without antibiotic treatment alters the protein composition of EVs and bMVs, we carried out proteomic profiling on both fractions using the antibiotic concentrations that produced the strongest effects in our flow cytometry assays. We also included total plasma proteomics to provide an overall view of systemic protein changes. However, the protein yields from the bMV fraction and from total plasma were too low for reliable pairwise comparisons (Supplemental Fig.\u0026nbsp;2A\u0026ndash;B). For this reason, we focused all downstream analyses on the EV fraction. sPLS-DA plotting showed clear separation between the MOCK control and the SA-infected/antibiotic-treated groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Furthermore, DEFA3, LTF, \u003cem\u003eSA\u003c/em\u003e ribosomal protein rplU, ITGB2, RAB7A, S100A8, UBA52, PTPRC, TF, ITGAM, SERPINA1, MYADM, SERPINC1, IQGAP2, ORM1, A1BG, SDCBP, C1QB, HPX, and TNFAIP8 contributed most to the variance in component 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). A volcano plot showed that DEFA3, LTF, rplU, MYADM, SDCBP, ITGB2, UBA52, PTPRC, S100A8, S100A9, ITGAM, RAB7A, and ITGA6 were significantly upregulated in the SA infection group compared to the MOCK control (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). A Venn diagram demonstrated that DEFA3, LTF, ITGB2, RAB7A, S100A8, UBA52, PTPRC, ITGAM, MYADM, and SDCBP were commonly identified in both the sPLS-DA and volcano plot analyses (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). Enrichment analysis revealed that these 11 proteins were significantly associated with pathways related to positive regulation of secretion, EV biogenesis, macrophage activation, bacterial infection, and host defense responses (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE). These proteomic results further support our EV flow cytometry findings (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). However, pairwise comparisons did not identify any significantly upregulated proteins in the piperacillin-tazobactam group compared to the SA-only group. In contrast, AZU1, EPB41, CAMP, and IGKV2-30 were significantly downregulated in the piperacillin-tazobactam group (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF). Interestingly, \u003cem\u003eSA\u003c/em\u003e proteins RplU and InfC were significantly increased in the vancomycin-treated group relative to the SA-only group (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG), supporting the EV flow cytometry data that indicated enhanced loading of \u003cem\u003eSA\u003c/em\u003e-derived proteins into human EVs following vancomycin treatment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). In the comparison between the moxifloxacin group and the \u003cem\u003eSA\u003c/em\u003e-only group, C1QB, PRKACB, SH3BGRL3, and OAZ1 were upregulated, whereas AZU1, APOD, IGKV2-30, and FCN2 were downregulated. Enrichment analysis did not reveal any significantly associated human pathways, consistent with the broadly similar proteomic profiles observed across the three antibiotic treatment groups compared to \u003cem\u003eSA\u003c/em\u003e infection alone.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSince EV\u0026ndash;Miltenyi bead flow cytometry and EV proteomics yielded the most reproducible results in \u003cem\u003eSA\u003c/em\u003e\u0026ndash;spiked plasma, we next investigated whether the human proteome is similarly altered in sepsis patients and whether bacterial ribosomal proteins can be detected in EVs. Prior to proteomic analysis, EV particle size and number were consistent across HC, BCN, and BCP groups (Supplemental Fig.\u0026nbsp;3A and B). Following data preprocessing steps including quality control, intensity normalization, and reproducibility-based filtering across biological replicates, a total of 424 proteins were confidently identified across samples from HC (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6), BCN (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6), and BCP (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6) groups. sPLS-DA analysis revealed clear separation between the HC group and the infection groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Notably, the overall proteomic profile of BCP patients also showed distinct separation from that of BCN patients, indicating further stratification based on blood culture status (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Differential analysis identified 57 significantly enriched proteins in the BCN group and 69 in the HC group (fold change\u0026thinsp;\u0026gt;\u0026thinsp;2) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Among the top five uniquely enriched proteins in BCN patients were SAA1, ANXA1, PLSCR1, SERPINA3, and APCS. Pathway enrichment analysis of the BCN-associated proteins revealed significant involvement in blood microparticles, positive regulation of exocytosis, defense response to bacteria, and immune-related pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). In the BCP group, 36 proteins were significantly enriched compared to HC, while 38 were enriched in HC (fold change\u0026thinsp;\u0026gt;\u0026thinsp;2) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). The top five uniquely enriched proteins in BCP patients were CRP, SAA2, SERPINA1, SERPINA3, and ANXA2. Enrichment analysis of BCP-associated proteins revealed similar involvement in blood microparticles, exocytosis, bacterial defense, and immune pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). Notably, the \u003cem\u003eSA\u003c/em\u003e infection pathway was also significantly enriched in BCP patients relative to HC (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). The observed enrichment of pathways related to blood microparticles and exocytosis in both BCN and BCP patients, as revealed by proteomic analysis, further supports our EV flow cytometry findings (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). To further explore host responses in culture-stratified infection states, we compared the proteomic profiles of BCN and BCP patients directly. Ten proteins were significantly enriched in each group (Supplemental Fig.\u0026nbsp;3C). Proteins enriched in BCN patients, including the metabolic enzyme LDHA; immune modulators ANXA1, ANXA3, ANXA5, and ANXA6; the oxidative burst enzyme CYBB; inflammatory mediator S100A9; complement-associated CLU; eosinophil granule protein PRG2; and antimicrobial peptides DEFA1/DEFA3, were associated with bacterial defense responses, endocytic vesicle formation, and leukocyte chemotaxis (Supplemental Fig.\u0026nbsp;3D). In contrast, proteins enriched in BCP patients, including immunoglobulin variants, ARHGDIB, NAPA, UNC13D, DSTN, HBB, and ESAM, were linked to immune recognition, vesicle trafficking, cytoskeletal remodeling, and vascular function, highlighting distinct host responses in blood culture\u0026ndash;positive sepsis (Supplemental Fig.\u0026nbsp;3C). Additionally, 24 proteins were commonly enriched in both BCN and BCP groups, based on a Venn diagram analysis of significant proteins across the three pairwise group comparisons (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF). To assess global proteomic changes associated with infection, we compared the combined infection group (Bac_infection_group: BCN\u0026thinsp;+\u0026thinsp;BCP, n\u0026thinsp;=\u0026thinsp;12) to HC (n\u0026thinsp;=\u0026thinsp;6). This analysis identified 57 proteins significantly enriched in the infection group and 70 enriched in HC (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG). Enrichment analysis of the infection-associated proteins again highlighted key pathways including blood microparticles, positive regulation of exocytosis, \u003cem\u003eSA\u003c/em\u003e infection, secretory granule membrane, innate immune activation, complement activation, and chemotaxis (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eH). Together, these findings indicate that EV proteomes from sepsis patients, whether blood culture\u0026ndash;positive or \u0026ndash;negative, are characterized by enhanced activity in vesicle-mediated transport, innate immune activation, and host defense pathways when compared to HC.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study presents a multi-tiered investigation into how EVs reflect host responses to bacterial infection and antibiotic treatment, combining an \u003cem\u003ein vitro SA\u003c/em\u003e infection model with EV profiling from patients with culture-positive and culture-negative sepsis. Using a combination of EV flow cytometry, bead-based isolation, and high-resolution label-free proteomics, we demonstrate that the EV proteome dynamically changes in response to the bacterial challenge and that these changes are modulated by specific antibiotic treatments. Notably, we identified distinct proteomic signatures in culture-positive and culture-negative sepsis patients that share common immune activation pathways but also exhibit stratifying features based on blood culture status.\u003c/p\u003e\n\u003cp\u003eOur \u003cem\u003eex vivo\u003c/em\u003e data demonstrated that \u003cem\u003eSA\u003c/em\u003e infection leads to a significant increase in CD45⁺ host-derived EVs carrying bacterial antigens (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eB), indicating active vesicle-mediated host\u0026ndash;pathogen interaction. This signal was further amplified by vancomycin treatment, which correlated with increased detection of \u003cem\u003eSA\u003c/em\u003e ribosomal proteins such as RplU and InfC in the EV fraction (Figs. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eB and \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eG). These results suggest that vancomycin may facilitate the incorporation of bacterial components into host EVs, possibly through enhanced bacterial lysis or altered vesicle trafficking pathways. Notably, fluorescence nanoparticle tracking analysis (F-NTA) showed no significant differences in the total number of EVs across experimental groups (Figs. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eD\u0026ndash;F), indicating that changes in vesicle content, rather than vesicle abundance, reflect the host response to infection and antibiotic treatment.\u003c/p\u003e\n\u003cp\u003eIn our \u003cem\u003eSA\u003c/em\u003e model, vancomycin and piperacillin-tazobactam, which target the bacterial cell wall and induce membrane stress, enhanced the incorporation of bacterial proteins into host EVs, whereas moxifloxacin, which acts on intracellular enzymes without damaging the membrane, showed no such effect, highlighting the importance of cell wall disruption in driving vesicle-associated protein release. This observation aligns with growing evidence that bactericidal antibiotics, including \u0026beta;-lactams and glycopeptides, can stimulate the release of bMVs as part of a general stress response\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. These vesicles often carry cytoplasmic and periplasmic proteins, nucleic acids, and misfolded proteins, contributing to bacterial survival, antibiotic resistance, and immune modulation\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. In gram-negative bacteria, early electron microscopy studies demonstrated that polymyxin treatment in \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e and \u003cem\u003eEscherichia coli\u003c/em\u003e (\u003cem\u003eE.coli\u003c/em\u003e) markedly increased membrane vesicle production\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Moreover, bMVs protect \u003cem\u003eE. coli\u003c/em\u003e and enterotoxigenic \u003cem\u003eE. coli\u003c/em\u003e (\u003cem\u003eETEC\u003c/em\u003e) from polymyxin B, colistin, and T4 bacteriophage by adsorbing these agents. Treatment with these antibiotics induces a substantial increase in bMV output, as measured by bMV protein content normalized to the number of CFU (colony forming unit)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eAlthough \u003cem\u003eSA\u003c/em\u003e lacks an outer membrane, our findings suggest that similar stress-responsive vesicle dynamics may occur in this gram-positive context, particularly in response to bactericidal antibiotic treatment. We used the DSM 20231 strain, a type strain from clonal complex 8 that carries multiple methicillin resistance genes, reflecting an intrinsic potential for \u0026beta;-lactam resistance. Consistent with prior work by Kim et al.\u003csup\u003e33\u003c/sup\u003e, we observed that vancomycin treatment enhanced the incorporation of bacterial ribosomal proteins such as RplU and InfC into host-derived EVs, likely due to increased bacterial lysis or altered processing of bacterial components. However, our study differs from Kim\u0026rsquo;s in two important ways. First, whereas Kim et al.\u003csup\u003e33\u003c/sup\u003e used pure bacterial cultures and directly isolated bMVs, our approach involved spiking \u003cem\u003eSA\u003c/em\u003e into human plasma and isolating host-derived EVs for proteomic analysis. Second, while Kim\u0026rsquo;s study identified bacterial \u0026beta;-lactamases and penicillin-binding proteins within MRSA-derived EVs that conferred antibiotic protection, we did not detect such resistance enzymes in the investigated EVs. This difference likely reflects the dominance of host proteins in our plasma-based samples, which can obscure lower-abundance bacterial proteins. Nevertheless, we were able to detect \u003cem\u003eSA\u003c/em\u003e ribosomal proteins within host EVs: RplU was significantly elevated in the \u003cem\u003eSA\u003c/em\u003e\u0026ndash;only group compared to the MOCK control, and both RplU and InfC showed further increases following vancomycin treatment. These findings indicate that bacterial components are successfully incorporated into host EVs in response to infection and antibiotic stress. Similar effects were also observed with ampicillin, a \u0026beta;-lactam that, like vancomycin, targets the peptidoglycan layer and induces membrane destabilization\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Although their molecular targets differ, both antibiotics appear to promote bacterial protein release through cell wall damage, facilitating their entry into host EVs. These observations are in line with Andreoni et al.\u003csup\u003e35\u003c/sup\u003e, who showed that certain \u0026beta;-lactam antibiotics such as flucloxacillin and ceftaroline enhanced vesicle formation in \u003cem\u003eSA\u003c/em\u003e through peptidoglycan weakening, and with Kim et al.\u003csup\u003e36\u003c/sup\u003e, who demonstrated that sub-inhibitory concentrations of antibiotics triggered MV release in \u003cem\u003eEnterococcus faecium\u003c/em\u003e and enhanced pro-inflammatory host responses. Together, these results support a broader role for bactericidal antibiotics in shaping EV content during host\u0026ndash;pathogen interaction.\u003c/p\u003e\n\u003cp\u003eIn line with growing evidence that EV reflect immune dynamics in systemic infection, our study demonstrates that both BCP and BCN sepsis patients exhibit robust changes in the EV proteome compared to HC. Notably, we observed substantial enrichment of innate immune mediators, complement components, and acute-phase proteins in plasma EVs, consistent with prior findings summarized in Chang Tian et al.\u003csup\u003e37\u003c/sup\u003e and other studies. Notably, several proteins known to mediate inflammation, immune cell recruitment, and opsonization were elevated, including CRP, SAA1, SAA2, SAA4, SERPINA1, SERPINA3, APCS, ANXA1, and CD177. Complement-related proteins such as C1QA, C1R, C1S, and SERPING1 were also enriched, along with coagulation and vascular injury markers like FGG, VWF, and ITIH3. Additional upregulated proteins such as PLSCR1, SDCBP, HP, ORM1, AGT, CTSG, MFGE8, CP, and SLC25A3 further support active EV-mediated immune signaling and endothelial responses (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). These protein signatures reinforce the concept that EVs act as carriers of immune and inflammatory mediators in sepsis, offering insight into both shared and context-specific pathways of disease progression. CRP, a known common sepsis marker\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, \u003csup\u003e39\u003c/sup\u003ewas also detected in our EV proteome, consistent with findings by Yan Xu et al.\u003csup\u003e39\u003c/sup\u003e. However, we did not detect Serine palmitoyltransferase 3 (SPTLC3), possibly due to differences in EV isolation methods, as their study used ultracentrifugation with density gradients, while we employed antibody bead\u0026ndash;based capture.\u003c/p\u003e\n\u003cp\u003eChanhee Park et al.\u003csup\u003e40\u003c/sup\u003e and Murao et al.\u003csup\u003e13\u003c/sup\u003e both identified key EV proteins linked to sepsis, several of which overlap with our findings. These include CRP, SERPIN (A1, A3, and G1), ORM1, VMW, SAA (1, 2, and 4), complements (C1R, C1QA, and C1S), and ITIH3, all of which were enriched in the sepsis EV proteome of this study. Notably, SERPINA1 and SERPINA3 are inhibitors of neutrophil serine proteases, playing dual roles in controlling inflammation during sepsis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. In parallel, our identification of C1QA, C1R, C1S, and SERPING1 underscores activation of the classical complement pathway, reinforcing its critical role in driving proinflammatory and prothrombotic responses in sepsis. These complement components contribute not only to pathogen recognition and clearance but also to endothelial activation, immune cell modulation, and the regulation of T-cell responses, further supporting the central involvement of the complement system in EV-mediated immune dysregulation during sepsis\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eInterestingly, several proteins identified in our dataset, such as SDCBP (syntenin), PLSCR1, CP, HP, AGT, FGG, APCS, CTSG, SLC25A3, MFGE8, and CD177, were not reported in their analyses. This suggests that our antibody bead-based EV isolation method may reveal additional or less-characterized components of the host response in sepsis. Notably, SDCBP (syntenin) was overexpressed in both the BCN and BCP groups compared to HC (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). This observation is consistent with our flow cytometry analysis using EVs isolated by Miltenyi beads (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e), as well as with our previous publication\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e, where western blot analysis also showed an increasing trend of SDCBP expression in sepsis patients relative to healthy controls. However, despite these changes in protein expression, the particle numbers remained unchanged in both our previous study and the current dataset (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). Together, these findings suggest that sepsis primarily alters the molecular composition of EVs rather than their overall abundance.\u003c/p\u003e\n\u003cp\u003eOur findings support that PLSCR1, which is overexpressed in sepsis patients compared to HC, plays a key role in protecting lung epithelial cells from \u003cem\u003eSA\u003c/em\u003e \u0026alpha;-toxin\u0026ndash;induced cell death by mediating IFN\u0026alpha;-driven resistance mechanisms\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. Ceruloplasmin (CP), a classic acute-phase protein, has recently been implicated in sepsis-induced cardiotoxicity through its involvement in cuproptosis, a newly identified form of regulated cell death\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Haptoglobin (HP) is another well-established acute-phase protein that mitigates hemoglobin-driven oxidative stress\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. Angiotensinogen (AGT) shows a variable but inducible response to inflammatory stimuli, suggesting context-dependent regulation during acute stress\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. Fibrinogen gamma chain (FGG), particularly its \u0026gamma;\u0026prime; variant, contributes to host defense in \u003cem\u003eSA\u003c/em\u003e sepsis by modulating bacterial aggregation and enhancing survival outcomes\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. SLC25A3, a mitochondrial carrier involved in transporting solutes like citrate and succinate, is associated with metabolic reprogramming during inflammation, although its specific role in sepsis remains unclear\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. MFGE8, previously reported to be downregulated in septic patients and protective against oxidative stress and ferroptosis, was found to be elevated in EVs from the BCN and BCP groups in our study; however, the functional significance of EV-associated MFGE8 in sepsis remains to be elucidated\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. CD177, a neutrophil activation marker involved in chemotaxis, was significantly upregulated in BCN and BCP compared to healthy controls (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e), consistent with previous studies showing elevated CD177 expression in circulating neutrophils during septic shock\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e and increased levels of CD177⁺ neutrophil-derived microvesicles in patients with bacteremia\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn conclusion, our study validates many established immune-modulating components of EVs described in the literature while adding novel evidence for underappreciated proteins in both culture-positive and culture-negative sepsis patients. The enrichment of antimicrobial peptides, annexins, Reactive oxygen species-generating enzymes, and vesicle trafficking proteins in patient-derived EVs expands the current understanding of EV-mediated immune regulation. These findings support the utility of EV profiling as a tool for uncovering mechanistic differences between sepsis subtypes and for identifying candidate biomarkers relevant to early diagnosis, pathogen-independent host response, and targeted therapy.\u003c/p\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003cbr\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe R scripts used for nanoparticle tracking analysis (NTA), extracellular vesicle (EV) flow cytometry, and proteomics data analysis, together with the imputed and normalized ProteinGroups files generated by MaxQuant and Perseus, have been uploaded to GitHub.\u003c/p\u003e\n\u003cp\u003ehttps://doi.org/10.5281/zenodo.17961832\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Faculty of Medicine at Ludwig-Maximilians-Universität München (Ethikkommission bei der Medizinischen Fakultät der LMU München) under Protocol #19-872. Written informed consent was obtained from all subjects prior to the collection of human blood samples. In addition to primary samples, only commercially available established cell lines were used for this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author(s) declare financial support was received for the research, authorship, and/or publication of this article. This project was funded by the German Federal Ministry for Economic Affairs and Climate Action by the “Central Innovation Program for small and medium-sized enterprises (SMEs)” (ZIM) under grant numbers KK5368301AD1, KK5367301AD1, and KK5022312AD1. The TEM images of the present study were supported by the “pro patient” grant (Grant No. pp18-08) and Krebsliga 462 beider Basel (Grant No. 18-463 2016). The Orbitrap Fusion Lumos mass spectrometer was funded in part by the German Research Foundation (INST 95/1436-1 \u0026nbsp;FUGG).\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eDC: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft. MP: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – review \u0026amp; editing. GS: Funding acquisition, Project administration, Writing – review \u0026amp; editing. AM: Funding acquisition, Project administration, Writing – review \u0026amp; editing. FB: Project administration, Writing – review \u0026amp; editing. CL : Data curation, Software, Writing – review \u0026amp; editing. SW: Data curation, Software, Writing – review \u0026amp; editing. BK: Writing – review \u0026amp; editing. MY: Writing – review \u0026amp; editing. CZ: Writing – review \u0026amp; editing. LM: Funding acquisition, Project administration, Resources, Writing – review \u0026amp; editing. MR: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – review \u0026amp; editing.\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eThe authors would like to thank Franziska Hackbarth for her excellent technical assistance and maintenance of mass spectrometers.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eRudd, K. E. et al. Global, regional, and national sepsis incidence and mortality, 1990\u0026ndash;2017: analysis for the Global Burden of Disease Study. \u003cem\u003eLancet\u003c/em\u003e \u003cb\u003e395\u003c/b\u003e, 200\u0026ndash;211. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0140-6736(19)32989-7\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(19)32989-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhnuma, T. et al. Epidemiology, Resistance Profiles, and Outcomes of Bloodstream Infections in Community-Onset Sepsis in the United States. \u003cem\u003eCrit. Care Med.\u003c/em\u003e \u003cb\u003e51\u003c/b\u003e, 1148\u0026ndash;1158. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/CCM.0000000000005870\u003c/span\u003e\u003cspan address=\"10.1097/CCM.0000000000005870\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao, M., Wang, G. \u0026amp; Xie, J. Immune dysregulation in sepsis: experiences, lessons and perspectives. \u003cem\u003eCell. Death Discov\u003c/em\u003e. \u003cb\u003e9\u003c/b\u003e, 465. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41420-023-01766-7\u003c/span\u003e\u003cspan address=\"10.1038/s41420-023-01766-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKwiecinski, J. M. \u0026amp; Horswill, A. R. Staphylococcus aureus bloodstream infections: pathogenesis and regulatory mechanisms. \u003cem\u003eCurr. Opin. Microbiol.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e, 51\u0026ndash;60. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.mib.2020.02.005\u003c/span\u003e\u003cspan address=\"10.1016/j.mib.2020.02.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUrbonaviciute, V. et al. Factors masking HMGB1 in human serum and plasma. \u003cem\u003eJ. Leukoc. Biol.\u003c/em\u003e \u003cb\u003e81\u003c/b\u003e, 67\u0026ndash;74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1189/jlb.0306196\u003c/span\u003e\u003cspan address=\"10.1189/jlb.0306196\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThavarajah, T. et al. The plasma peptides of sepsis. \u003cem\u003eClin. Proteom.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e, 26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12014-020-09288-5\u003c/span\u003e\u003cspan address=\"10.1186/s12014-020-09288-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaigo, K. et al. The proteomic profile of circulating pentraxin 3 (PTX3) complex in sepsis demonstrates the interaction with azurocidin 1 and other components of neutrophil extracellular traps. \u003cem\u003eMol Cell Proteomics\u003c/em\u003e 11, M111 015073 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/mcp.M111.015073\u003c/span\u003e\u003cspan address=\"10.1074/mcp.M111.015073\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWasserman, A. et al. Septic patients presenting with apparently normal C-reactive protein: A point of caution for the ER physician. \u003cem\u003eMed. (Baltim).\u003c/em\u003e \u003cb\u003e98\u003c/b\u003e, e13989. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/MD.0000000000013989\u003c/span\u003e\u003cspan address=\"10.1097/MD.0000000000013989\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYanez-Mo, M. et al. Biological properties of extracellular vesicles and their physiological functions. \u003cem\u003eJ. Extracell. Vesicles\u003c/em\u003e. \u003cb\u003e4\u003c/b\u003e, 27066. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3402/jev.v4.27066\u003c/span\u003e\u003cspan address=\"10.3402/jev.v4.27066\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeber, B., Henrich, D., Hildebrand, F., Marzi, I. \u0026amp; Leppik, L. The Roles of Extracellular Vesicles in Sepsis and Systemic Inflammatory Response Syndrome. \u003cem\u003eShock\u003c/em\u003e 59, 161\u0026ndash;172 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/SHK.0000000000002010\u003c/span\u003e\u003cspan address=\"10.1097/SHK.0000000000002010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Langhe, N. et al. Mapping bacterial extracellular vesicle research: insights, best practices and knowledge gaps. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 9410. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-024-53279-1\u003c/span\u003e\u003cspan address=\"10.1038/s41467-024-53279-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYou, B., Yang, Y., Wei, J., Zhou, C. \u0026amp; Dong, S. Pathogenic and therapeutic roles of extracellular vesicles in sepsis. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 1535427. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fimmu.2025.1535427\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2025.1535427\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorris, D. C. et al. Proteomic Profiles of Exosomes of Septic Patients Presenting to the Emergency Department Compared to Healthy Controls. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/jcm9092930\u003c/span\u003e\u003cspan address=\"10.3390/jcm9092930\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003edo Nascimento, M. F. et al. Circulating extracellular vesicles as potential biomarkers and mediators of acute respiratory distress syndrome in sepsis. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 5512. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-025-89783-7\u003c/span\u003e\u003cspan address=\"10.1038/s41598-025-89783-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEvans, L. et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021. \u003cem\u003eCrit. Care Med.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, e1063\u0026ndash;e1143. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/CCM.0000000000005337\u003c/span\u003e\u003cspan address=\"10.1097/CCM.0000000000005337\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenecke, L., Chiang, D. M., Ebnoether, E., Pfaffl, M. W. \u0026amp; Muller, L. Isolation and analysis of tumor\u0026ndash;derived extracellular vesicles from head and neck squamous cell carcinoma plasma by galectin\u0026ndash;based glycan recognition particles. \u003cem\u003eInt. J. Oncol.\u003c/em\u003e \u003cb\u003e61\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3892/ijo.2022.5423\u003c/span\u003e\u003cspan address=\"10.3892/ijo.2022.5423\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShevchenko, A., Tomas, H., Havli, J., Olsen, J. V. \u0026amp; Mann, M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. \u003cem\u003eNat. Protoc.\u003c/em\u003e \u003cb\u003e1\u003c/b\u003e, 2856\u0026ndash;2860. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nprot.2006.468\u003c/span\u003e\u003cspan address=\"10.1038/nprot.2006.468\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuchner, K., Ludwig, C., Kerpes, R. \u0026amp; Becker, T. Higher induction temperatures and the native secretion signal peptide promote rye prolamin 75k gamma-secalin production in Komagataella phaffii. \u003cem\u003eMicrob. Cell. Fact.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, 185. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12934-025-02809-7\u003c/span\u003e\u003cspan address=\"10.1186/s12934-025-02809-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, M. S. C. et al. The proteome of bacterial membrane vesicles in Escherichia coli-a time course comparison study in two different media. \u003cem\u003eFront. Microbiol.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 1361270. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmicb.2024.1361270\u003c/span\u003e\u003cspan address=\"10.3389/fmicb.2024.1361270\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTyanova, S., Temu, T. \u0026amp; Cox, J. The MaxQuant computational platform for mass spectrometry-based shotgun proteomics. \u003cem\u003eNat. Protoc.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 2301\u0026ndash;2319. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nprot.2016.136\u003c/span\u003e\u003cspan address=\"10.1038/nprot.2016.136\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTyanova, S. et al. The Perseus computational platform for comprehensive analysis of (prote)omics data. \u003cem\u003eNat. Methods\u003c/em\u003e. \u003cb\u003e13\u003c/b\u003e, 731\u0026ndash;740. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nmeth.3901\u003c/span\u003e\u003cspan address=\"10.1038/nmeth.3901\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRStudio Integrated Development Environment for R v. 2025.05.0\u0026thinsp;+\u0026thinsp;496 (RStudio, PBC, Boston, MA).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR: A language and environment for statistical computing v. 4.5.1. R Foundation for Statistical Computing, Vienna, Austria, (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, G. Thirteen years of clusterProfiler. \u003cem\u003eInnov. (Camb)\u003c/em\u003e. \u003cb\u003e5\u003c/b\u003e, 100722. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.xinn.2024.100722\u003c/span\u003e\u003cspan address=\"10.1016/j.xinn.2024.100722\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, S. et al. Using clusterProfiler to characterize multiomics data. \u003cem\u003eNat. Protoc.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 3292\u0026ndash;3320. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41596-024-01020-z\u003c/span\u003e\u003cspan address=\"10.1038/s41596-024-01020-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu, T. et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. \u003cem\u003eInnov. (Camb)\u003c/em\u003e. \u003cb\u003e2\u003c/b\u003e, 100141. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.xinn.2021.100141\u003c/span\u003e\u003cspan address=\"10.1016/j.xinn.2021.100141\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, G., Wang, L. G., Han, Y. \u0026amp; He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. \u003cem\u003eOMICS\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 284\u0026ndash;287. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1089/omi.2011.0118\u003c/span\u003e\u003cspan address=\"10.1089/omi.2011.0118\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, G. \u0026amp; He, Q. Y. ReactomePA: an R/Bioconductor package for reactome pathway analysis and visualization. \u003cem\u003eMol. Biosyst\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, 477\u0026ndash;479. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1039/c5mb00663e\u003c/span\u003e\u003cspan address=\"10.1039/c5mb00663e\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRohart, F., Gautier, B., Singh, A., KA, L. C. \u0026amp; mixOmics An R package for 'omics feature selection and multiple data integration. \u003cem\u003ePLoS Comput. Biol.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, e1005752. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pcbi.1005752\u003c/span\u003e\u003cspan address=\"10.1371/journal.pcbi.1005752\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMacNair, C. R. \u0026amp; Tan, M. W. The role of bacterial membrane vesicles in antibiotic resistance. \u003cem\u003eAnn. N Y Acad. Sci.\u003c/em\u003e \u003cb\u003e1519\u003c/b\u003e, 63\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/nyas.14932\u003c/span\u003e\u003cspan address=\"10.1111/nyas.14932\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoike, M., Iida, K. \u0026amp; Matsuo, T. Electron microscopic studies on mode of action of polymyxin. \u003cem\u003eJ. Bacteriol.\u003c/em\u003e \u003cb\u003e97\u003c/b\u003e, 448\u0026ndash;452. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1128/jb.97.1.448-452.1969\u003c/span\u003e\u003cspan address=\"10.1128/jb.97.1.448-452.1969\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1969).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManning, A. J. \u0026amp; Kuehn, M. J. Contribution of bacterial outer membrane vesicles to innate bacterial defense. \u003cem\u003eBMC Microbiol.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 258. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/1471-2180-11-258\u003c/span\u003e\u003cspan address=\"10.1186/1471-2180-11-258\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, S. W. et al. Significant increase in the secretion of extracellular vesicles and antibiotics resistance from methicillin-resistant Staphylococcus aureus induced by ampicillin stress. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 21066. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-020-78121-8\u003c/span\u003e\u003cspan address=\"10.1038/s41598-020-78121-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUddin, M. J. et al. The Role of Bacterial Membrane Vesicles in the Dissemination of Antibiotic Resistance and as Promising Carriers for Therapeutic Agent Delivery. \u003cem\u003eMicroorganisms\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/microorganisms8050670\u003c/span\u003e\u003cspan address=\"10.3390/microorganisms8050670\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndreoni, F. et al. Antibiotics Stimulate Formation of Vesicles in Staphylococcus aureus in both Phage-Dependent and -Independent Fashions and via Different Routes. \u003cem\u003eAntimicrob. Agents Chemother.\u003c/em\u003e \u003cb\u003e63\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1128/AAC.01439-18\u003c/span\u003e\u003cspan address=\"10.1128/AAC.01439-18\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, M. H. et al. Production of Membrane Vesicles by Enterococcus faecium Cultured With or Without Subinhibitory Concentrations of Antibiotics and Their Pathological Effects on Epithelial Cells. \u003cem\u003eFront. Cell. Infect. Microbiol.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fcimb.2019.00295\u003c/span\u003e\u003cspan address=\"10.3389/fcimb.2019.00295\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian, C. et al. Extracellular vesicles participate in the pathogenesis of sepsis. \u003cem\u003eFront. Cell. Infect. Microbiol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 1018692. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fcimb.2022.1018692\u003c/span\u003e\u003cspan address=\"10.3389/fcimb.2022.1018692\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePepys, M. B. \u0026amp; Hirschfield, G. M. C-reactive protein: a critical update. \u003cem\u003eJ. Clin. Invest.\u003c/em\u003e \u003cb\u003e111\u003c/b\u003e, 1805\u0026ndash;1812. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1172/JCI18921\u003c/span\u003e\u003cspan address=\"10.1172/JCI18921\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, Y. et al. Exosomal proteome analysis of human plasma to monitor sepsis progression. \u003cem\u003eBiochem. Biophys. Res. Commun.\u003c/em\u003e \u003cb\u003e499\u003c/b\u003e, 856\u0026ndash;861. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.bbrc.2018.04.006\u003c/span\u003e\u003cspan address=\"10.1016/j.bbrc.2018.04.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark, C. et al. Proteomic profiling of plasma extracellular vesicles identifies signatures of innate immunity, coagulation, and endothelial activation in septic patients. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 21871. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-025-06430-x\u003c/span\u003e\u003cspan address=\"10.1038/s41598-025-06430-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, Y. X. \u0026amp; Li, L. Identification of potential biomarkers of sepsis using bioinformatics analysis. \u003cem\u003eExp. Ther. Med.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 1689\u0026ndash;1696. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3892/etm.2017.4178\u003c/span\u003e\u003cspan address=\"10.3892/etm.2017.4178\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZetoune, F. S. \u0026amp; Ward, P. A. Role of Complement and Histones in Sepsis. \u003cem\u003eFront. Med. (Lausanne)\u003c/em\u003e. \u003cb\u003e7\u003c/b\u003e, 616957. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fmed.2020.616957\u003c/span\u003e\u003cspan address=\"10.3389/fmed.2020.616957\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaarman, A., Milder, F., van Strijp, J. \u0026amp; Rooijakkers, S. Complement inhibition by gram-positive pathogens: molecular mechanisms and therapeutic implications. \u003cem\u003eJ. Mol. Med. (Berl)\u003c/em\u003e. \u003cb\u003e88\u003c/b\u003e, 115\u0026ndash;120. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00109-009-0572-y\u003c/span\u003e\u003cspan address=\"10.1007/s00109-009-0572-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBroker, B. M., Mrochen, D. \u0026amp; Peton, V. The T Cell Response to Staphylococcus aureus. \u003cem\u003ePathogens\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/pathogens5010031\u003c/span\u003e\u003cspan address=\"10.3390/pathogens5010031\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReithmair, M. et al. Cellular and extracellular miRNAs are blood-compartment-specific diagnostic targets in sepsis. \u003cem\u003eJ. Cell. Mol. Med.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 2403\u0026ndash;2411. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jcmm.13162\u003c/span\u003e\u003cspan address=\"10.1111/jcmm.13162\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLizak, M. \u0026amp; Yarovinsky, T. O. Phospholipid scramblase 1 mediates type i interferon-induced protection against staphylococcal alpha-toxin. \u003cem\u003eCell. Host Microbe\u003c/em\u003e. \u003cb\u003e11\u003c/b\u003e, 70\u0026ndash;80. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.chom.2011.12.004\u003c/span\u003e\u003cspan address=\"10.1016/j.chom.2011.12.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYan, J., Li, Z., Li, Y. \u0026amp; Zhang, Y. Sepsis induced cardiotoxicity by promoting cardiomyocyte cuproptosis. \u003cem\u003eBiochem. Biophys. Res. Commun.\u003c/em\u003e \u003cb\u003e690\u003c/b\u003e, 149245. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.bbrc.2023.149245\u003c/span\u003e\u003cspan address=\"10.1016/j.bbrc.2023.149245\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoh, T. et al. Fucosylated haptoglobin promotes inflammation via Mincle in sepsis: an observational study. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 1342. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-025-56524-3\u003c/span\u003e\u003cspan address=\"10.1038/s41467-025-56524-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoden, M., Klett, C., Hasmann, T. \u0026amp; Hackenthal, E. Angiotensinogen: an acute-phase protein? \u003cem\u003eHypertension\u003c/em\u003e 23, I126\u0026ndash;130 (1994). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1161/01.hyp.23.1_suppl.i126\u003c/span\u003e\u003cspan address=\"10.1161/01.hyp.23.1_suppl.i126\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNegron, O. et al. Fibrinogen gamma' promotes host survival during Staphylococcus aureus septicemia in mice. \u003cem\u003eJ. Thromb. Haemost\u003c/em\u003e. \u003cb\u003e21\u003c/b\u003e, 2277\u0026ndash;2290. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jtha.2023.03.019\u003c/span\u003e\u003cspan address=\"10.1016/j.jtha.2023.03.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIacobazzi, V. et al. Mitochondrial carriers in inflammation induced by bacterial endotoxin and cytokines. \u003cem\u003eBiol. Chem.\u003c/em\u003e \u003cb\u003e398\u003c/b\u003e, 303\u0026ndash;317. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1515/hsz-2016-0260\u003c/span\u003e\u003cspan address=\"10.1515/hsz-2016-0260\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu, W. et al. MFG-E8 has guiding significance for the prognosis and treatment of sepsis. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 20916. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-022-25601-8\u003c/span\u003e\u003cspan address=\"10.1038/s41598-022-25601-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDemaret, J. et al. Identification of CD177 as the most dysregulated parameter in a microarray study of purified neutrophils from septic shock patients. \u003cem\u003eImmunol. Lett.\u003c/em\u003e \u003cb\u003e178\u003c/b\u003e, 122\u0026ndash;130. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.imlet.2016.08.011\u003c/span\u003e\u003cspan address=\"10.1016/j.imlet.2016.08.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTimar, C. I. et al. Antibacterial effect of microvesicles released from human neutrophilic granulocytes. \u003cem\u003eBlood\u003c/em\u003e \u003cb\u003e121\u003c/b\u003e, 510\u0026ndash;518. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1182/blood-2012-05-431114\u003c/span\u003e\u003cspan address=\"10.1182/blood-2012-05-431114\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8384635/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8384635/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSepsis accounts for nearly 20% of global mortality, with antibiotic resistance worsening clinical outcomes. Rapid antibiotic administration and accurate pathogen identification remain crucial. It is well now known that extracellular vesicles (EVs) from human cells and bacterial membrane vesicles (bMVs) play a central role in the interaction between host and pathogen and represent promising biomarkers for early infections. This study investigated how antibiotic exposure alters EV responses in \u003cem\u003eStaphylococcus aureus\u003c/em\u003e (\u003cem\u003eSA\u003c/em\u003e)\u0026ndash;spiked blood and compared these findings with EV proteome profiles from bacteremia patients.\u003c/p\u003e \u003cp\u003eWhole blood from healthy donors was spiked with \u003cem\u003eSA\u003c/em\u003e (Multiplicity of infection: 0.001) and treated with piperacillin\u0026ndash;tazobactam (Pip-Tazo), vancomycin, or moxifloxacin at clinically relevant concentrations. EVs were isolated using the Miltenyi Pan EV Kit, and bMVs were captured with magnetic beads conjugated to anti-OmpA and anti-GroEL. EVs and bMVs were analyzed using bead-based flow cytometry. Proteomics of total plasma, EVs and bMVs high-resolution LC\u0026ndash;MS/MS. Patients\u0026rsquo; serum EVs from 6 healthy controls and 12 bacteremia patients (6 blood culture\u0026ndash;positive, 6 culture-negative) were processed using the same workflow to assess both host and bacterial proteins.\u003c/p\u003e \u003cp\u003eFlow cytometry revealed that levels of CellMask Orange⁺ (CMO⁺) CD45⁺ PanEV⁺ SA⁺ extracellular vesicles increased in blood samples exposed to low concentrations of Pip\u0026ndash;Tazo and high concentrations of vancomycin, despite minimal changes in vesicle size distribution and total particle counts. Proteomic analysis identified notable alterations in EV-associated proteins, including strong elevation of the ribosomal protein bL21 in SA-spiked samples and those treated with vancomycin. Gene ontology analysis indicated enrichment of innate immune and exosome-related pathways. In patient samples, EVs were enriched with acute-phase proteins such as PLSCR1, haptoglobin, CRP, and SAA1\u0026ndash;4, along with canonical EV markers CD81 and MFGE8, irrespective of culture positivity.\u003c/p\u003e \u003cp\u003eAntibiotic treatment leads to significant remodeling of the EV proteome, characterized by enhanced presence of immune and bacterial response proteins, even when vesicle numbers remain constant. These protein shifts appeared in both culture-positive and culture-negative patient samples, supporting the idea that EV-associated proteins could serve as early, host-derived indicators of bloodstream infection.\u003c/p\u003e","manuscriptTitle":"Host-Defense Extracellular Vesicle Protein Changes in Antibiotic- and Staphylococcus aureus–Treated Blood","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-27 07:29:26","doi":"10.21203/rs.3.rs-8384635/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-25T14:09:13+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-25T00:40:28+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-18T14:39:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"144399220695767118566938593426199253086","date":"2026-03-18T06:50:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"252217004433701896112448474987805076925","date":"2026-03-17T09:52:32+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"132682561871671787865189225980001819947","date":"2026-03-17T04:12:54+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"315650082022137385653104847814011851328","date":"2026-03-01T11:36:02+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-27T11:34:27+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-16T07:41:59+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-09T10:02:02+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-01-09T09:50:19+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2c869994-23cc-4b32-bd38-8ff11b886ac1","owner":[],"postedDate":"February 27th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":61945521,"name":"Health sciences/Diseases"},{"id":61945522,"name":"Biological sciences/Immunology"},{"id":61945523,"name":"Biological sciences/Microbiology"}],"tags":[],"updatedAt":"2026-05-18T13:53:19+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-27 07:29:26","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8384635","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8384635","identity":"rs-8384635","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}