Exploring the Proteomic Landscape and Immunomodulatory Functions of Edwardsiella piscicida Derived Extracellular Vesicles Authors: Mawalle Kankanamge Hasitha Madhawa Dias, E.H.T. Thulshan Jayathilaka, Jayasinghage Nirmani Chathurangika Jayasinghe, Nipuna Tennakoon, Chamilani Nikapitiya, Ilson Whang, Mahanama De Zoysa Published: 2025 ABSTRACT Extracellular vesicles (EVs) have garnered attention in research for their potential as biochemical transporters and immune modulators, crucial for regulating the host immune system. The present study was conducted to isolate and characterize EVs from Gram negative bacteria Edwardsiella piscicida (EpEVs) and investigate their proteomic profile and immune responses. Isolation of EpEVs was carried out using ultracentrifugation method. Transmission electron microscopy results confirmed the spherical shape of EpEVs. The average size and zeta potential were 85.3 ± 1.8 nm and -8.28 ± 0.41 mV, respectively. EpEVs consisted of 1,487 distinct proteins. Subcellular localization analysis revealed that “cell” and “cell part” were the most predominant areas for protein localization. Proteins associated with virulence, along with several chaperones that facilitate protein folding and stability, were also present. No toxicity was detected when EpEVs were treated to fathead minnow (FHM) cells up to 100 μg/ml. Fluorescent-labeled EpEVs showed cellular internalization in FHM cells at 24 h post treatment (hpt). In-vitro gene expression in Raw 264.7 cells showed upregulation of interleukin (Il)6, Il1β, and interferon (Ifn)β with simultaneous upregulation of anti-inflammatory Il10. In vivo, gene expression revealed that except for heat shock protein (hsp)70, all other genes were upregulated suggesting that EpEVs induced the expression of immune-related genes. Western blot analysis showed increased protein levels of tumor necrosis factor (Tnf)α in EpEVs-treated spleen tissue of zebrafish. Our results confirm that EpEVs can be successfully isolated using the ultracentrifugation method. Furthermore, exploring immunomodulatory mechanism of EpEVs is essential for their potential use as novel therapeutics in fish medicine. Introduction Extracellular vesicles (EVs) have emerged as a promising biomaterial in modern therapeutics, particularly in the development of immunotherapy and targeted drug-delivery systems [1]. Like eukaryotes, bacteria also produce EVs, known as bacterial EVs (BEVs), which remain a relatively underexplored area with vast untapped potential. BEVs can deliver both natural and bioengineered cargo [2]. These nanovesicles, typically ranging from 20-400 nm in diameter, are produced from both gram positive and gram negative bacteria [2]. These spherical-shaped vesicular bodies contain a diverse composition of molecules, including structural components like lipopolysaccharide (LPS) and peptidoglycan, toxins, periplasmic and cytoplasmic proteins, and nucleic acids [2-4]. Biogenesis of gram negative BEVs is mainly explained by two proposed models; outer membrane blebbing and explosive cell lysis [2]. In the outer membrane blebbing model, instability in the peptidoglycan layer causes the outer membrane to bud into the periplasm, while explosive cell lysis occurs when the inner membrane protrudes into the periplasm through weak points in the peptidoglycan layer [2]. Following the release, BEVs have the potential to be internalized by both prokaryotic and eukaryotic cells, where they can deliver their cargo to modulate complex biological processes and facilitate intercellular communication [5]. Both pathogenic and non-pathogenic bacteria can produce BEVs, with varying functions depending on their cargo, type of bacteria, and its target. These vesicle components play roles in various biological processes, including the expression of virulence, phage infection, cellular metabolites transportaiton, and horizontal gene transfer [2]. Furthermore, it has been shown that BEVs exhibit protein cargo sorting and selective loading [6]. This cargo selectivity is closely linked with the biogenesis process, indicating that it is a highly regulated and orchestrated biological phenomenon rather than a random occurrence [2]. For the selective cargo loading, electrostatic interactions, curvature-recognizing protein interactions, and lipoprotein interactions work tandemly during BEVs biogenesis [4]. Furthermore, research by Chen et al. in Shewanella vesiculosa proposed the “P49 protein secretion mechanism”, suggesting that foreign protein can be incorporated into BEVs via fusing with the P49 protein [7]. Therefore, understanding BEVs proteins and their interactions with host cells is crucial for discovering disease biomarkers, which can be applied in diagnostics, screening, prognosis, and treatment [8]. In-depth knowledge of these cargo molecules in specific bacteria could broaden our knowledge of their functional roles in relation to each host. Proteomics is a valuable approach that aids researchers in identifying and elucidating BEVs-host interactions, making it a powerful tool for disease diagnostics [9]. Although several studies exist on mammalian pathogenic BEVs, research on fish pathogenic BEVs remains a relatively unexplored subject area. Table 1 summarizes the available proteomic studies on fish pathogenic BEVs and their activities. Edwardsiella piscicida is a gram negative, rod-shaped bacterium that is a well-established pathogen in both freshwater and marine fish [15-17]. Additionally, E. piscicida poses a serious risk to human health as the consumption of contaminated food products can lead to cross-species transmission of the infection to humans [15]. Moreover, the pathogenesis of E. piscicida is largely driven by its Type III and Type VI secretion systems (T3SS and T6SS), which are crucial virulence factors and key targets for vaccine development against edwardsiellosis [15]. Although a single study has explored the potential of E. tarda EVs as a vaccine candidate, the present study provides a broader approach by focusing on isolation of BEVs from E. piscicida using the ultracentrifuge method, along with the characterization of bacterial marker proteins. Moreover, this study includes a comprehensive proteomic analysis of EpEVs, as well as assessments of their toxicity and immunomodulatory effects in Raw 264.7 cells and zebrafish. Materials and Methods E. piscicida Culture and Isolation of EVs A loopful of E. piscicida (lab strain) was streaked onto Brain Heart Infusion (BHI) plates supplemented with 1%NaCl and incubated overnight at 25°C. A single colony was picked and inoculated into BHI+1% NaCl broth (4 ml), followed by overnight incubation at 25°C with continuous agitation at 180 rpm. For EVs isolation, a large culture (1 L) of E. piscicida was prepared by adding an overnight-grown bacteria culture (OD595 ~ 1.0) into BHI+1% NaCl broth at 0.5% v/v and incubated for 24 h. Following the bacteria centrifugation, the supernatant was filtered (0.2 μm) and concentrated (10×) using 100 kDa Amicon® Ultra-15 centrifugal filters (Millipore, Cork, Ireland). The supernatant was ultracentrifuged using a CP100NX UC system (Koki Holdings Co. Ltd., Japan) equipped with a P28S2 swinging rotor (Himac, Ibaraki, Japan) at 100,000 ×g; 4°C for 3 h. The resulting EpEVs pellet was washed using ultracentrifuged and filtered phosphate-buffered saline (PBS) (Gibco-BRL, USA) following the same aforementioned conditions. Pelleted EpEVs were dissolved in filtered PBS (900 μL) and kept at -80°C until further study. Characterization of EpEVs Biogenesis, Morphology, Particle Size, Concentrations, and Surface Charge EpEVs biogenesis was examined with field emission scanning electron microscopy (FE-SEM) following a pre-described method [18]. Briefly, bacterial cell pellet from an overnight grown and reinoculated fresh E. piscicida culture was washed with PBS and dehydrated using a gradient ethanol series (30, 50, 70, 80, 90, and 100%) and coated with platinum. Finally, observation was done using FE-SEM (Serion FEI, Netherlands). The Morphology of EpEVs was evaluated using FE-transmission electron microscopy (FE-TEM) opting for a Tecnai G2 F30 super-twin; FEI system (Hillsboro, USA) according to a method described previously [19]. Particle size and distribution were assessed by nanoparticle tracking analysis (NTA) using a NanoSight NS300 system (Malvern Technologies, UK) while the surface charge of EpEVs was measured by zeta potential analyzer (Zetasizer Nano ZSP, UK) according to the method described by Nikapitiya et al. [19]. Protein Profile and Molecular Markers of EpEVs The protein content of isolated EpEVs was quantified using the Bradford assay (Bio-Rad, USA) according to the manufacturer’s instructions. For both in vitro and in vivo treatments, EpEVs were quantified based on the protein content obtained from this assay. The protein profile of EpEVs was determined using 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described previously [20]. EpEVs sample (25 μg) were mixed with 4 × Laemmli sample buffer (Bio Rad) and denatured at 95°C for 10 min. The sample was loaded onto Mini-PROTEAN TGX gels (Bio Rad) and electrophoresed at 100 mV for 90 min. Protein bands were visualized by Coomassie blue staining (Biosesang, Republic of Korea). For Western blot analysis, EpEVs proteins were separated on an SDS-PAGE gel and transferred to Immobilon-P Polyvinylidene difluoride membranes (Millipore Corporation, USA). Membranes were blocked with 5% (w/v) bovine serum albumin (BSA) prepared in Tris-buffered saline with 0.05% Tween 20 (TBST) (LPS Solution, Republic of Korea) at 26 ± 2°C for 1 h. They were then incubated overnight at 4°C with primary antibodies against flagellin (Abcam, Republic of Korea) and E. coli outer membrane protein A (OmpA) (Antibody Research Corporation, USA). After washing with PBS containing 0.05% Tween 20 (PBST) (three times), membranes were incubated with secondary antibodies (Table 2) at 26 ± 2°C for 1 h. Protein bands were visualized using the Western Femto ECL Kit (LPS solution, Republic of Korea) and detected with a chemiluminescence imaging system (Fusion Solo S, Vilber, France). Protein Preparation and HPLC-MS/MS Based Proteomic Analysis Proteomic analysis of EpEVs was conducted as described in our previous publication [21]. Briefly, EpEVs samples were treated with a reducing agent dithiothreitol (DTT; 10 mM) at 37°C for 30 min followed by incubation with iodoacetamide (IAM; 55 mM) at 25°C for 45 min in dark. After processing, samples were centrifuged (25,000 ×g at 4°C for 15 min) and quantified using the Bradford assay (Bio-Rad), and subjected to electrophoresis for protein separation. During the enzymatic digestion process, 100 μg of protein solution was mixed with 50 mM ammonium bicarbonate (NH4HCO3) to make final SDS concentration below 0.1%. Trypsin (2.5 μg) was added at a protein-to-enzyme ratio of 20:1 and incubated at 37°C for 4 h for digestion. The resulting peptides were desalted using a Strata X column (10 mg/ml, Tubes; Phenomenex, USA) and dried under vacuum using a MaxiVac beta freeze dryer (LaboGene™, Bjarkesvej, Denmark). Peptide separation was carried out using a Shimadzu (LC-20AB) high-performance liquid chromatography (HPLC) system equipped with a Gemini C18 column (4.6 mm internal diameter, 5 μm column size, 25 cm column length). Separation was performed at a flow rate of 1 ml/min and 10 fractions were collected and subsequently lyophilized. EpEVs were analyzed for peptide sequencing and quantification using Orbitrap Exploris 480 mass spectrometer system (Thermo Fisher Scientific, USA) coupled to an UltiMate 3000 UHPLC system (Dionex, USA). The separated peptides were ionized using a nano-electrospray ionization (ESI) source and analyzed in data-dependent acquisition (DDA) mode. Peptide samples were reconstituted, centrifuged, and separated using a defined gradient at a flow rate of 500 nL/min. MS parameters were optimized to ensure effective peptide sequencing and protein quantification, including ion source voltage, scanning range, resolution, collision energy, and precursor peptide ion selection criteria. Bioinformatics of Proteomic Data In the DDA library construction, Mascot software (version 2.3.02) was employed to identify peptides from protein sequences in an Edwardsiella-specific database. Trypsin was employed for protein cleavage with allowance for up to one missed cut. The peptide-protein matching error threshold was set at 1% and the minimum acceptable peptide length was seven amino acids. Fixed modifications included carbamidomethyl at cysteine sites, while variable modifications included methionine oxidation, conversion of glutamine to pyroglutamate, and deamidation of pyroglutamate. To identify protein groups, tandem MS results were pre-processed and re-scored using Percolator software to improve the matching accuracy. The data were filtered with a false discovery rate (FDR) of 1% at the spectral level, yielding a list of significant spectra and peptides. Gene ontology (GO) annotation was performed using the DAVID database (https://david.ncifcrf.gov/) accessed on 2023.12.20 under default settings to assess cluster enrichment. Pathway enrichment analysis was conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG), while protein-protein interaction (PPI) predictions of EpEVs and E. tarda were conducted using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (Version 12.0; https://string-db.org) at default settings (accessed on 2024.04.26). These analyses provided valuable insights into the functional roles and interaction networks of the profiled proteins in the BEVs. Assessment of In Vitro and In Vivo Toxicity of EpEVs Fathead minnow (FHM) epithelial cells and Raw 264.7 murine macrophage cells were used to determine the toxicity and safety dose of EpEVs. FHM cells were cultured in sealed cap culture flasks at 25°C, using L15 medium supplemented with 10% fetal bovine serum (FBS) (Wellgene Co., Ltd., Republic of Korea). Raw 264.7 cells were maintained in a Dulbecco's Modified Eagle Medium (DMEM) (Wellgene Co., Ltd., Republic of Korea) supplemented with 10% FBS and 1% antibiotic mixture (Penicillin/Streptomycin) (Gibco, USA). The cytotoxicity of EpEVs was assessed using the Cellrix Viability Assay Kit (MediFab, Republic of Korea) following the manufacturer's instructions. FHM and Raw 264.7 cells were seeded in a 96-well plate at 2 × 105 cells/ml (100 μl/well) and incubated at 25°C and 37°C for 12 h, respectively. EpEVs (10 μl; concentrations ranging from 0 to 100 μg/ml) were treated to the cells and incubated for 24 h, after which Cellrix reagent (10 μl per well) was introduced. After incubation at 37°C for 2.5 h, the optical density at 450 nm (OD450) was measured using a microplate reader (Bio Rad, Japan). Based on the results, the safety dose was determined. Reactive oxygen species (ROS) generation was measured in seeded Raw 264.7 cells following a 24 h incubation. The cells were treated with 2'7'dichlorodihydro-fluorescein diacetate (DCFHDA) (Sigma-Aldrich, USA) at a concentration of 5 μg/ml for 30 min. Excess dye was removed by washing with PBS and ROS generation was observed under a fluorescence microscope (DMi8; Leica, Germany). Adult zebrafish were reared, bred, and embryos were collected following established procedures [22]. Healthy larvae [60 h post-fertilization (hpf)] were selected for the experiment and treated with EpEVs (10-100 μg/ml; 10 larvae per treatment) for 96 h. Mortality was monitored every 12 h. After 96 h of treatment, five larvae from each group were incubated with DCFHDA (5 μg/ml) for 30 min to assess EpEVs-induced ROS generation, following a previously published method [22]. Excess dye was removed by washing with PBS, and imaging was performed using a Leica S8 APO stereo microscope (Nikon SMZ100, Japan) equipped with a SFA fluorescent filter (NIGHTSEA, USA). Fluorescent Labeling and Internalization of EpEVs in FHM Cells Fluorescent labeling of EpEVs was conducted to confirm cellular uptake using fish cells. The procedure followed the guidelines provided by the ExoSparkler kit (Dojindo Molecular Technologies Inc., USA) for the fluorescent labeling of EpEVs. Briefly, EpEVs (10 μg) were diluted in PBS and treated with Mem dye-green solution (2 μl). The mixture was incubated for 30 min, then transferred to a filtration tube and centrifuged at 3,000 ×g at 26 ± 2°C for 5 min. The samples were washed twice with 100 μl of PBS each time, followed by re-centrifugation under the same conditions. Finally, the labeled EpEVs were reconstituted in PBS. Pre-seeded FHM cells (1.25 × 104 cells/ml) in 12 well plates (1 ml/well) were treated with fluorescent-labeled EpEVs (20 μg in 50 μl) for 24 h. Subsequently, the cells were washed with PBS and stained with 4',6-diamidino-2-phenylindole (DAPI; 300 nM for 5 min) (Sigma-Aldrich). EpEVs internalization was then observed using a fluorescence microscope. Analysis of In Vitro Immunomodulatory Activity of EpEVs in Raw 264.7 Cells Raw 264.7 cells (1 × 105 cells/well) were seeded in 6-well plates and incubated at 37°C. After 12 h, the cells were treated with EpEVs at two different concentrations (5 and 10 μg/ml) for 24 h, after which they were harvested. Total RNA was isolated using the NucleoSpin RNA Mini Kit (Macherey-Nagel, Germany). RNA concentrations were measured using a NanoDrop One spectrometer (Thermo Fisher Scientific), and cDNA was synthesized using the PrimeScript 1st strand cDNA Synthesis Kit (TaKaRa, Japan). Gene-specific primers for quantitative real-time PCR (qRT-PCR) are in Table 3. The cDNA was diluted to 10 ng/ml and qRT-PCR was performed using the Thermal Cycler Dice Real-Time System (TaKaRa). The relative mRNA expression of each gene was analyzed using the 2−ΔΔCT method with Gapdh serving as the housekeeping gene [23]. Analysis of Immune Gene and Protein Expression in EpEVs-Treated Zebrafish Adult zebrafish (average body weight of 0.45±0.05g) were acclimatized for two weeks before being intraperitoneally injected with EpEVs (5 and 10 μg/fish), following a protocol in a previous publication [23]. Kidney and spleen tissues (3 kidneys per replicate) from each group were collected at 6 and 24 h post-treatment (hpt) and stored at -80°C. For immune gene expression analysis, RNA was extracted from kidney tissues and cDNA was synthesized according to the manufacturer's protocols. Zebrafish gene-specific primers are listed in Table 3, with β-actin used as the housekeeping gene. For protein expression analysis, spleen tissue was used. Tissues were lysed using ProEX CeTi Lysis buffer (TransLab, Republic of Korea) and the total protein content was quantified using the Bradford method. After preparing the samples, 25 μg of protein from each sample was loaded into a 12% SDS-PAGE gel, and immunoblotting was performed. Details of used primary antibodies and corresponding secondary antibodies are provided in Table 2. Finally, the membranes were incubated and visualized using a chemiluminescence detection system. Statistical Analysis The significance between each group was determined using one-way analysis of variance (ANOVA) with GraphPad Prism software version 8 (USA). Mean separation was conducted using an unpaired two-tailed t-test, and statistical significance was defined as p < 0.05. Results Isolation and Characterization of EpEVs Initially, we conducted molecular identification and FE-SEM analysis to confirm that the selected bacterium was E. piscicida. 16S rRNA gene sequencing results were analyzed using the EzBioCloud 16S database (CJ Bioscience, Inc., Republic of Korea) and showed 100% similarity (as of 2023.03.10) to the 16S rRNA genes of known E. piscicida. This, in turn, aids in isolating EVs specifically from E. piscicida. FE-SEM imaging displayed a prominent rod-shaped bacterium resembling the morphology of E. piscicida (Fig. 1A). When examining E. piscicida under higher magnification (×100K), spherical structures were observed that attached to the outer membrane of the bacterial surface, providing insights into the biogenesis of EpEVs (Fig. 1B). Using FE-TEM imaging, EpEVs were identified as small, spherical bilayer membrane structures (Fig. 1C). NTA-based size distribution analysis of EpEVs revealed that a larger proportion of particles falls within the 50-150 nm size range (Fig. 1D). Moreover, the visual representation of particles showed evenly distributed EpEVs, in the sample (Fig. 1d). Percentile values showed that 90% of the particles were smaller than 128.5 ± 3.7 nm, while only 10% were smaller than 55.5 ± 0.9 nm (data not shown). Furthermore, the mean particle size was 85.3 ± 1.8 nm (Fig. 1E). The concentrations of EpEVs were 2.36 ± 0.11 × 1012 particles/ml, while the protein recovery was 2 mg/l of bacterial culture. Zeta potential analysis further revealed a negative surface charge of EpEVs (-8.28 ± 0.41 mV). Total Protein Profile and Molecular Characterization of EpEVs Understanding the total protein profile and molecular characteristics of BEVs is essential for elucidating their biological functions. The total protein profile of EpEVs was examined using Coomassie blue-stained SDS-PAGE, highlighting the major protein components and their relative abundances (Fig. 2A). Six prominent protein bands were observed in the range of 10-180 kDa, with the highest band intensity localized between the molecular mass range of 35-48 kDa. Western blotting was conducted to detect the presence of BEVs marker proteins (Fig. 2B). Flagellin and OmpA were expressed at approximately 47 and 37.5 kDa, respectively and the intensity of the flagellin band was greater than that of OmpA. This finding was further supported by EpEVs proteomic analysis which showed higher intensity-based absolute quantification (iBAQ) values for flagellin isomers compared to OmpA isoforms (Table S1). Proteomic Analysis of EpEVs We conducted a proteomic analysis of EpEVs to study the entire proteome using HPLC-MS/MS. Raw mass spectrometry data were searched against the Mascot MS/MS database for peptide identifications and resulted in a total spectra of 56,179 (Table S2). Next, a subset of the total spectra was denoted as identified spectra of 15,911, which were matched with known spectra. This resulted in 10,745 identified peptides including unique peptides of 9,157 (85.22%). After searching the Edwardsiella_qc database (6,957 total sequences), we identified 1,487 proteins from EpEVs. The most abundant protein mass range was 20-30 kDa (353 proteins) whereas the least number of proteins (31) had a molecular mass range of 90-100 kDa (Fig. 2C). In peptide length distribution analysis, the highest spike was noted for the 11-15 peptide length range (3110 peptides) (Fig. 2D). In the peptide length range of 20 – infinity (∞), only 13.85% of peptides are represented, where the highest was observed between 21-25 (831). Unique peptide distribution revealed about 51.64% of the identified proteins had three or more peptides while about 73.50% of the identified proteins had 2 or more peptides (Fig. 2E). About 71.35% and 51.38% of peptides showed more than 10 and 20% sequence coverage, respectively (Fig. 2F) which suggested that a higher amount of peptides had a lower percentile coverage. To find out the physiological contexts, interactions, and regulatory roles of EpEVs, subcellular localization of identified proteins was predicted. As expected the majority of protein localization was classified under ‘cellular component’ (30.94%). Furthermore, almost similar protein numbers (30.41%) represented the ‘cell part’ which together comprised 61.35% of the total proteins (Fig. 3A). ‘Extracellular region’ (0.01%) was the least number of proteins present in EpEVs. Furthermore, GO analysis was conducted to map the molecular functions for the identified proteins categorizing them into three groups namely, biological processes, cellular components, and molecular functions (Fig. 3B). The majority of the proteins belonged to molecular functions where ‘catalytic activity’ related proteins such as Chorismate mutase, Cytochrome c-type protein, Glutathione synthetase, etc. were highly abundant (905). Proteins related to ‘binding’ (HTH-type transcriptional regulator IscR, Ribosomal RNA small subunit methyltransferase B, etc.) were the second highest out of the category (722). ‘Cell and cell part ’-related proteins belonging to the cellular component category were 586 and 576, respectively. Biological processes regulating proteins are annotated into 19 functional groups and out of them ‘metabolic processes’ [Ribonuclease G, 23S rRNA (guanosine-2'-O-)-methyltransferase RlmB, etc.] and ‘cellular processes’ (Signal peptide peptidase SppA, Macrodomain Ter protein, etc.) regulating proteins were the top two categories (760 and 722, respectively) while only one protein (Aldehyde-alcohol dehydrogenase) was recorded for ‘carbon utilization’ which was shown to be the least. Furthermore, from the identified proteins various virulence factors, such as proteins related to Type III secretion system (EscV/YscV/HrcV family type III secretion system export apparatus protein, TyeA family type III secretion system gatekeeper subunit, and CesD/SycD/LcrH family type III secretion system chaperone), flagellin and Omps and their assemblies, multiple chaperone proteins (HtrA protease/chaperone protein, Chaperone protein DnaJ, Heat shock chaperone IbpB), etc. were also present (Table 4). From the observed results it is apparent that the identified EpEVs proteins have a vast array of functional properties that could be a considerable advantage for using it as an immunomodulating agent or as a vaccine candidate. PPIs of EpEVs were determined using the STRING database by subjecting 1487 of identified EpEVs proteins. Protein networking was mapped using the default settings of STRING where the host organism was selected as E. tarda (Fig. 4A). Out of the identified proteins, 673 interactions were predicted with E. tarda proteins showing their association and involvement with various biological and functional processes. Fig. 4B shows three of the identified key protein interactions important for protein export, microbial metabolism in various environments, and metabolic pathways, further suggesting EpEVs potential for cargo transportation and metabolism. KEGG pathway enrichment analysis revealed that 12 pathways were enriched (Table S3), out of which both aminoacyl-tRNA biosynthesis and RNA degradation pathways had 100% of the observed gene count (16/16 and 8/8, respectively) in relation to the background genes. Furthermore, from the 12 enriched pathways five were related to metabolic processes which show insights into the potential metabolic activity of EpEVs proteins. In Vitro and In Vivo Toxicity Analysis and Cellular Internalization of EpEVs The in vitro toxicity of EpEVs was assessed using FHM and Raw 264.7 cells to understand the cell-specific toxicity of EpEVs and to determine the safe dose for further experiments. At low concentrations of EpEVs (below 10 μg/ml), Raw 264.7 cells did not show adverse effects (Fig. 5A). However, cells treated at 25 μg/ml showed a reduction of cell viability up to 52.85% with clear morphological changes (data not shown). Moreover, the viability of EpEVs-treated Raw 264.7 cells was reduced drastically above 25 μg/ml. The IC50 value of EpEVs was 17.34 μg/ml for Raw 264.7 cells. Next, we compared the toxicity of EpEVs for FHM cells. Interestingly, EpEVs did not demonstrate any apparent toxic effect even at 100 μg/ml (Fig. 5B) and we confirm IC50 value of EpEVs for FHM cells could be above 100 μg/ml. Furthermore, intracellular ROS generation was analyzed in Raw 264.7 cells (Fig. 5C). Cellular ROS level was significantly increased (p<0.05) after 5 μg/ml EpEVs treatment (no statistical significance for 10 μg/ml of EpEVs) up to 100 μg/ml showing green fluorescent in cells compared to the negative control. Interestingly, EpEVs treated cells exhibited less ROS generation compared to the positive control (5 μM; H2O2). Based on overall toxicity results safety dose for in vitro studies was confirmed as 5 and 10 μg/ml of EpEVs for Raw 264.7 cells. For the determination of in vivo toxicity, we determined the survival % of EpEVs-treated zebrafish larvae at 60 hpf. No mortality was observed in EpEVs-treated larvae (100 μg/ml) at 96 hpt (Data not shown). Additionally, we investigated whether EpEVs treatment could induce oxidative stress in zebrafish larvae by analyzing in vivo ROS level (Fig. 5D). Comparatively low level of green fluorescence was detected only above 50 μg/ml treatment of EpEVs and it was also lower than that of H2O2 (positive control) treated larvae group. We used fluorescent-labeled EpEVs to confirm cellular internalization into FHM cells (Fig. 5E). A preliminary experiment was conducted to optimize the peak time of internalization, and cells showed the highest fluorescence intensity at 24 hpt. The blue fluorescence in FHM cells is related to the DAPI-stained nuclei and the green fluorescence in FHM cells corresponds to internalized EpEVs. Compared to the control, florescent labeled EpEVs-treated FHM cells showed slight green fluorescence suggesting the successful internalization of EpEVs at 24 hpt. In Vitro Immunomodulatory Activity of EpEVs To study whether EpEVs have immunomodulatory effects, we used EpEVs-treated Raw 264.7 cells and investigated the transcriptional regulation by qRT-PCR at 6 and 24 hpt (Fig. 6). Among the selected Toll-like receptors (Tlrs), only Tlr2 was significantly upregulated (p < 0.05) upon both EpEVs treatments (5 μg/ml; 1.83*-fold, and 10 μg/ml; 1.55*-fold) at 6 hpt. Myeloid differentiation primary response 88 (Myd88) was upregulated only at 5 μg/ml treatment of EpEVs (6 hpt). Expression of other Tlrs such as Tlr4 and Tlr5 expressed almost basal levels except for Tlr4 (6 hpt) and Tlr5 (24 hpt) at 10 μg/ml. Pro and anti-inflammatory molecules including Il1β, Il6, and Il10 were significantly upregulated (p < 0.05) in both concentrations and time-dependent manner. Moreover, Il6 had the highest expression level (1088.52-fold at 10 μg/ml; 24 hpt). Il1β had the second highest relative fold expression at 24 hpt (5 μg/ml; 544.42*-fold and 10 μg/ml; 379.27*-folds). Ifnα was upregulated for both concentrations (5 μg/ml; 3.12-fold and 10 μg/ml; 2.82-fold) only at 24 hpt. Ifnβ expression levels were over 30-fold upregulated in both EpEVs concentration and time points. Genes related to antioxidant enzymes including catalase (Cat) and superoxide dismutase1 (Sod1) were evaluated and results confirm the upregulated Cat expression at 24 hpt in both concentrations (5 μg/ml; 2.35*-fold, and 10 μg/ml; 2.24*-fold) while Sod1 remained at a basal level for all time points. In comparison, Il6 was the only gene that showed a concentration-dependent transcriptional upregulation at both time points. In Vivo Immunomodulatory Activity of EpEVs In vivo, immunomodulatory responses of EpEVs were evaluated by isolating mRNA from EpEVs-treated adult zebrafish kidney tissue by qRT-PCR (Fig. 7A). Overall, all the selected genes were upregulated in zebrafish with 5μg/fish of EpEVs at both 6 and 24 hpt. However, at 6 hpt, 5 μg/fish EpEVs treatment did not upregulate most genes (tlr2, tlr4, myd88, il10, and tnfα) which were upregulated in the 10 μg/fish EpEVs-treated zebrafish. Additionally, tlr5b showed upregulation in both EpEVs-treated groups. The myd88 was upregulated only in the 10 μg/fish EpEVs treatment group. At 24 hpt, expression of all tlr genes and myd88 was increased. The highest fold change in tlr expression was observed for tlr4 in the 10 μg/fish EpEVs-treated group at 24 hpt. Both il1β and il8 showed upregulation in response to EpEVs treatment at both 6 and 24 hpt. Specifically, il1β exhibited the highest relative fold expression (30.99-fold) at 6 hpt in the 10 μg/fish treatment group. Antioxidant gene sod1 remained at basal level at 6 hpt but showed upregulation at 24 hpt. In addition, in vivo immunomodulation was evaluated using Western blotting analysis with spleen samples from adult zebrafish (Fig. 7B). Hsp90 levels were similarly expressed in both EpEVs-treated groups at both time points. Alp levels were induced at 6 hpt with a reduction observed at 24 hpt. Tnfα expression decreased in the 10 μg/fish EpEVs-treated group at 6 hpt and concentration-dependent reduction was observed at 24 hpt. Discussion BEVs released by pathogenic bacteria could potentially be developed as vaccine candidates due to their ability to present bacterial-specific virulence factors, which can trigger immune responses in the host [10]. In this study, we applied the ultracentrifugation technique to isolate BEVs from E. piscicida as it is one of the least disruptive methods for isolating BEVs, according to Libardo et al. [24]. We confirmed the host bacterium E. piscicida shares 100% similarity with the known E. piscicida strain (accession no. JRGQ0100013) isolated by Camus et al.[25]. Next, we confirmed that EpEVs are nanosized (85.3 ± 1.8 nm), spherical-shaped particles with clear margins, within the range described for BEVs (20-400 nm) [2]. The findings of Brudal et al. demonstrated that BEVs isolated from fish pathogenic bacterium Francisella noatunensis had an average size of 72.34 nm, which is relatively similar to our results [26]. However, Park et al. reported smaller-sized E. tarda EVs with an average diameter ranging from 10-40 nm [10]. The size difference may be attributed to the ultracentrifugation-based sucrose density gradient method used in the study. The major advantage of ultracentrifugation is its ability to yield a high quantity of EV extracellular vesicle particles with considerable purity. The negative charge of EVs is an overlooked characteristic that plays a vital role in their attachment to host cells and the internalization of cargo into EVs. According to Midekessa et al., the behavior of nanoscale molecules such as EVs, is highly influenced by their surface charge. Furthermore, the surface charge of EVs depends on factors such as the ionization level of membrane surface groups (e.g., COO- group of certain proteins), protonation, the presence of polymeric chains attached to the surface and H-bonds, types of intramolecular bonds, and the availability of ions in the solution [27]. Kaddour et al. reported that human blood-derived EVs (zeta-potential; −1.4 mV) were less efficiently internalized into human cervical epithelial cells, primary blood-derived monocytes, and primary peripheral blood lymphocytes compared to semen EVs (zeta-potential; −8.82 mV) [28]. Moreover, Kaddour et al. emphasized that the surface charge of EVs plays a crucial role in determining their attachment and internalization to target cells. Typically, gram negative BEVs exhibit a more pronounced negative charge compared to gram positive BEVs, primarily due to the presence of negatively charged phospholipids in the bacterial membrane of the respective strains [29]. However, the presence of cations such as Mg2+ and Ca2+ can stabilize the net charge, allowing BEVs to be stored long-term at -80°C without any damage [30]. Therefore, we proposed that EpEVs could serve as a promising biomaterial to use as a vaccine candidate or drug delivery agent in fish medicine. NTA further confirms the efficiency of ultracentrifugation, showing a relatively higher particle concentration of 2.36 ± 0.11 × 1012 particles/ml and a protein recovery of 2.0 mg/l from the bacterial supernatant. Won et al. reported that culturing a large batch (200 L) of Escherichia coli yielded 357 mg (1.79 mg/l) of total protein amount of BEVs, equivalent to 3.93 × 1015 particles (1.96 × 1010 particles/mL of starting culture) [31]. In contrast, its final concentration was lower compared to our results. Therefore, it is important to explore methods for isolating BEVs at larger scales to enhance isolation efficiency while reducing costs. Proteomic characterization of EpEVs was conducted to gain deeper insights into their structure and function, as well as to understand how each component may contribute to the overall functionality of EpEVs as a potential immunomodulatory agent or a vaccine candidate. Park et al. identified 74 proteins from outer membrane vesicles released by E. tarda, whereas the present study identified 1,487 proteins [10]. Out of the total E. piscicida proteome, this represents approximately 21.37%, indicating a significantly higher protein count and suggesting the presence of various heterogeneous proteins with diverse functions. Although it is the only available literature with similarities to EpEVs, the isolation procedure and proteomic analysis methods used were different from those in the current study, which may explain the differing results. EpEVs contain proteins such as hemolysin, Omps, and flagellin, Type III secretory system (T3SS) proteins, and chaperones all of which are essential for the pathogenesis of E. piscicida [32, 33]. Moreover, among the mentioned virulence factors, flagellin, and T3SS translocon protein were two of the five most abundant proteins identified in the total protein sample (data not shown). Results of the present study identified three T3SS proteins, two flagellin derivatives, nine Omps, their assembly factors, and 15 chaperone candidates. These components could be targeted in future studies to develop EpEVs as a vaccine for controlling Edwardsiellosis. PPIs were predicted using the STRING database to construct a network of interactions between EpEVs proteins and those of E. tarda. Furthermore, KEGG analysis revealed 12 enriched pathways in the predicted PPI network, with pathways related to various metabolic activities being highly enriched. Pathways such as aminoacyl-tRNA biosynthesis, protein export, and ribosome are known for their roles in protein synthesis and transportation [34]. Additionally, pathways related to purine metabolism, carbon metabolism, oxidative phosphorylation, metabolic pathways, biosynthesis of secondary metabolites, and microbial metabolism in diverse environments were identified as key pathways linked to various metabolic activities in the predicted PPIs. Together we provide important insights into how EpEVs inherit and carry a substantial portion of the proteins from their originating organism, as well as how these proteins interact in relation to protein biosynthesis, metabolism, and transportation which are key processes to consider when evaluating EpEVs as a vaccine candidate. However, it is important to note that further studies should focus on host PPIs with EpEVs and their underlying mechanisms. To further confirm the safety of EpEVs as therapeutic agents, both in vitro and in vivo studies were conducted. The toxicity of EpEVs was evaluated by treating Raw 264.7 murine macrophage cells, and FHM cells in vitro as well as zebrafish larvae in vivo. Results revealed IC50 of 17.34 μg/ml for EpEVs in Raw 264.7 cells, while FHM cells showed no cytotoxicity even at a concentration of 100 μg/ml. In contrast, Vanaja et al. explained that the primary cause of toxicity in BEVs is the presence of LPS [35]. They further reported that E. coli-derived EVs (125 μg/ml) caused more than 50% cell death in bone marrow-derived mast cells, accompanied by an increase in cytosolic LPS levels. This phenomenon could explain the slight increase in toxicity observed when EpEVs were treated to Raw 264.7 cells at higher concentrations. Furthermore, Matsuda et al. observed a 60% survival rate in zebrafish embryos at 96 hpt when treated with milk-derived EVs (2 × 1012 particles/ml) which is a relatively higher concentration than that used in the present study [36]. Excessive production of ROS can be detrimental to cells and tissues causing oxidative stress that damages the structural integrity and functionality of living organisms [37]. Therefore, measuring ROS generation provides a direct indication of the toxicity level of a sample when applied to cells and animals. According to the findings of Marzoog et al., E. coli-derived BEVs induced higher levels of ROS generation in a concentration-dependent manner (12.5-50 μg/ml) when treated with HT-29 human colon cancer cells, demonstrating their anti-cancer effect [38]. Overall, both in vitro and in vivo ROS generation showed no significant increase following EpEVs treatment, further supporting the conclusion that isolated EVs exhibit lower toxicity. Therefore, the safety dose of EpEVs was determined as ≤ 10 μg/ml for Raw 264.7 cells and ≤ 100 μg/ml for zebrafish. As EVs become increasingly important in drug delivery for specialized therapies, cellular internalization of EVs plays a crucial role in transporting the specified cargo to the target organ [5]. Cellular internalization of BEVs occurs through one of five identified mechanisms; endocytosis, membrane fusion, receptor-mediated signaling, phagocytosis, and micropinocytosis [39]. Membrane fusion and endocytosis facilitate EVs to enter host cells by forming early endosomes, which release their cargo in a process that is heavily dependent on EV size [39, 40]. In receptor-mediated signaling, host cell receptors, such as pattern recognition receptors (Tlrs) are activated to mediate host cellular signaling for BEVs internalization, a process currently utilized in cancer therapy [39]. In this study, fluorescent labeling of EpEVs was conducted by labeling the lipid bilayer, which is important for investigating their uptake by cells, biodistribution, and effect on cellular activities [41]. Additionally, it allows for the accurate characterization of individual EVs. The observed results confirm that EpEVs can be fluorescently labeled and they are successfully internalized into FHM cells. This suggests their potential as a delivery system. Further modifications could optimize the loading of specific drugs into EpEVs, as well as their release both in vitro and in vivo. Although the fundamental basis of pathogenic bacteria is to infect the host without triggering its defense mechanisms, certain structural components of the bacteria can activate the host immune system to control the infection. Since BEVs have a membrane structure similar to that of parent bacteria, these molecules can stimulate the host immune system without causing infection [6]. Flagellin, LPS, and peptidoglycan are key bacterial membrane components commonly found in BEVs. These molecules act as virulence factors, mimicking the presence of bacteria and being recognized by pathogen recognition receptors (PRRs) [Tlrs, nucleotide-binding oligomerization domain-like receptors (NLR), C-type lectin receptors, and RIG-1 like receptors (RLR)] [42, 43]. This is consistent with our in vitro and in vivo transcriptional analysis results. While Raw 264.7 cells showed upregulation of only Tlr2 (at 6 hpt), upregulation of tlr2, tlr4, and tlr5b in zebrafish was observed for 24 hpt for both concentrations suggesting that EpEVs have the ability to stimulate host Tlrs. Generally, the stimulation of host PRRs activates various transcription factors, which regulate cell signaling pathways, mediating the host’s innate immune responses and triggering the release of a range of pro-inflammatory (Tnf-α, Il1β, Il6, and Il8) and anti-inflammatory (Il10) cytokines [43]. These cytokines further activate immune cells, priming and maturing them to produce molecules that amplify the inflammatory responses, while also stimulating the production of type 1 Ifns, antimicrobial peptides, and chemokines that help control or mitigate the ongoing threat to the host and provide protection [43, 44]. The results of the present study show reveal that at 24 hpt both concentrations of EpEVs (5 and 10 μg/ml) induced upregulation for pro and anti-inflammatory cytokines in zebrafish and Raw 264.7 cells consistent with previous reports. Sivanantham et al. have demonstrated that E. coli-derived EVs (0.25 μg/ml) have activated Tlr4, which led to the release of Tnfα and Il1β pro-inflammatory cytokines in murine bone marrow-derived macrophage cells [42]. Moreover, a study conducted by Lagos et al. shows that EVs from F. noatunensis subsp. orientalis increased the expression of pro-inflammatory cytokines tnfα, il1β, and ifnγ which aligns with our findings [45]. In addition, a study conducted by Park et al. demonstrated the elevation of Tlr2, Il1β, Il6, TNFα, and IFNγ upon treatment of E. tarda (at present E. piscicida)-derived EVs at 0.2 μg/g of body weight when injected to olive flounders [10]. When compared to the treatment doses (~10 and 20 μg/g of body weight of fish) of the present study, the expression level was relatively lower, which could be due to other factors such as the use of different fish species, environmental factors, stress conditions as well as the age. Alp is a group of enzymes that catalyze the hydrolytic removal of phosphate ions from various molecules [46]. Specifically, they play a role in the detoxification of LPS derived from the outer membrane of gram-negative bacteria, thereby modulating host-bacterial interactions [46]. Hsp90 is a universally expressed protein, except in archaea, and plays a crucial role in regulating cell survival [47]. A study conducted by Bates et al. revealed that intestinal Alp-deficient zebrafish were hypersensitive to LPS and exhibited elevated expression of myd88 and tnfγ, suggesting an active inflammatory response [48]. Therefore, since Alp protein levels were comparatively high at 6 hpt, it can be assumed that Alps contributed to reducing the inflammatory response, while Hsp90 is important for cell survival and stress response. Furthermore, Cecil et al. reported that BEVs released from Porphyromonas gingivalis can trigger an initial pro-inflammatory response characterized by the production of TNFα and Il1β, which primes monocytes and macrophages. A subsequent exposure, however, induces the release of the anti-inflammatory Il10, promoting tolerance to endotoxins [49]. This finding may further support the results observed in the present study, which demonstrated the simultaneous expression of both pro-inflammatory (Il1β, Il6) and anti-inflammatory (Il6) cytokines. Cat and Sods are essential enzymes that activate inherent anti-oxidative mechanisms to minimize oxidative stress caused by stimulated inflammation [50]. Even though ROS production was relatively low when EpEVs were treated with Raw 264.7 cells and zebrafish larvae, oxidative stress resulting from immune response at 24 hpt was evident, as indicated by the high expression of Cat in Raw 264.7 cells and sod1 in zebrafish kidney tissues. Although cellular Sod1 remained at the basal level, Cat expression was upregulated at 24 hpt, likely due to increased levels of peroxide ions in the form of H2O2. Cat primarily converts H2O2 into H2O while comparatively higher toxic superoxide free radicals are managed by Sods [51]. Based on this, we can hypothesize that treatment with EpEVs may increase the H2O2 in Raw 264.7 cells, leading to an oxidative stress response. Al-Nedawi et al. reported that Lactobacillus rhamnosus-derived EVs induced Il10 and heme-oxygenase-1 expression in dendritic cells [52]. This finding may be linked to our in vitro and in vivo observations, where Il10 upregulation, in combination with increased levels of Cat and Sod1, regulated both the inflammatory response and oxidative stress simultaneously. Conclusion This study demonstrates that ultracentrifugation is an effective and convenient technique for isolating high-purity BEVs from gram negative E. piscicida, preserving their characteristic properties. Through comprehensive proteomic analysis, we identified specific proteins consistently present in BEVs, providing valuable insight into their biological functions. Additionally, we identified several common proteins as cargo molecules in EpEVs which may play crucial roles in intercellular communications and the transfer of biological information. Our findings demonstrate that EpEVs have been uptaken by cells. However, further research is needed to investigate mechanisms underlying the internalization of EpEVs and the functional outcomes of their uptaken in various cellular contexts. Furthermore, EpEVs exhibited minimal toxicity when tested with Raw 264.7 cells and zebrafish larvae. The immunomodulatory effects of EpEVs were highly promising when treated to Raw 264.7 cells and zebrafish. Overall, these results have significant implications for the development of EpEVs-based therapeutic strategies, including vaccines and drug delivery systems in fish medicine. Supplemental Materials Supplementary data for this paper are available on-line only at http://jmb.or.kr.