Comprehensive characterization of extracellular vesicles produced by environmental (Neff) and clinical (T4) strains of Acanthamoeba castellanii

ABSTRACT We conducted a comprehensive comparative analysis of extracellular vesicles (EVs) from two Acanthamoeba castellanii strains, Neff (environmental) and T4 (clinical). Morphological analysis via transmission electron microscopy revealed slightly larger Neff EVs (average = 194.5 nm) compared to more polydisperse T4 EVs (average = 168.4 nm). Nanoparticle tracking analysis (NTA) and dynamic light scattering validated these differences. Proteomic analysis of the EVs identified 1,352 proteins, with 1,107 common, 161 exclusive in Neff, and 84 exclusively in T4 EVs. Gene ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) mapping revealed distinct molecular functions and biological processes and notably, the T4 EVs enrichment in serine proteases, aligned with its pathogenicity. Lipidomic analysis revealed a prevalence of unsaturated lipid species in Neff EVs, particularly triacylglycerols, phosphatidylethanolamines (PEs), and phosphatidylserine, while T4 EVs were enriched in diacylglycerols and diacylglyceryl trimethylhomoserine, phosphatidylcholine and less unsaturated PEs, suggesting differences in lipid metabolism and membrane permeability. Metabolomic analysis indicated Neff EVs enrichment in glycerolipid metabolism, glycolysis, and nucleotide synthesis, while T4 EVs, methionine metabolism. Furthermore, RNA-seq of EVs revealed differential transcript between the strains, with Neff EVs enriched in transcripts related to gluconeogenesis and translation, suggesting gene regulation and metabolic shift, while in the T4 EVs transcripts were associated with signal transduction and protein kinase activity, indicating rapid responses to environmental changes. In this novel study, data integration highlighted the differences in enzyme profiles, metabolic processes, and potential origins of EVs in the two strains shedding light on the diversity and complexity of A. castellanii EVs and having implications for understanding host-pathogen interactions and developing targeted interventions for Acanthamoeba-related diseases. IMPORTANCE A comprehensive and fully comparative analysis of extracellular vesicles (EVs) from two Acanthamoeba castellanii strains of distinct virulence, a Neff (environmental) and T4 (clinical), revealed striking differences in their morphology and protein, lipid, metabolites, and transcripts levels. Data integration highlighted the differences in enzyme profiles, metabolic processes, and potential distinct origin of EVs from both strains, shedding light on the diversity and complexity of A. castellanii EVs, with direct implications for understanding host-pathogen interactions, disease mechanisms, and developing new therapies for the clinical intervention of Acanthamoeba-related diseases.

amoeba species.Thus, a detailed elucidation of the composition of A. castellanii EVs, critical virulence transmission factors, is potentially important for identifying new drug targets and development of novel therapies.

Isolation of EVs of A. castellanii
The EVs of A. castellanii from both Neff and T4 strains were obtained as previously described, following the MISEV2018 procedures (47,48).Trophozoites were cultivated in 175 cm 2 culture flasks containing 100 mL PYG, with a starting inoculum of 2.5 × 10 5 amoebae/mL.After 48 h cultivation, about 1 L of cultures were centrifuged at 800 g for 10 min to pellet the trophozoites.The supernatants were additionally centrifuged at 15,000 g for 10 min.Sediments were discarded, and the new resulting supernatants were concentrated 20× by ultrafiltration in Amicon (cutoff = 10 kDa).The concentrated material was ultracentrifuged at 100,000 g for 1 h at 4°C (Beckman Optima LE-80K ultracentrifuge using the 70Ti Fixed-Angle titanium rotor) and washed with phosphatebuffered saline (PBS: 0.2 g/L KH 2 PO 4 .0.2 g/L KCl.2.4 g/L Na 2 HPO 4 and 8 g/L NaCl; pH 7.2) by sequential centrifugations and resuspensions to isolate the EVs of A. castellanii.Protein concentrations were measured using the bicinchoninic acid (BCA) method with bovine serum albumin as standard, following the manufacturer's instructions (Thermo Fisher).Sterol concentrations were determined using the Amplex Red Cholesterol Assay Kit (ThermoFisher).Samples were frozen at −20°C until the moment of analysis.

Morphological analysis of EVs from Neff and T4 strains of A. castellanii
A combination of methods was used to characterize the physical properties of EVs of A. castellanii.EVs from both strains underwent evaluations via transmission electron microscopy (TEM).For this, EV preparations were fixed for 2 h at RT with 2.5% glutaral dehyde, 4% paraformaldehyde, and 5 mM CaCl 2 in 0.1 M sodium cacodylate buffer (pH 7.2).EVs were subsequently washed with PBS, followed by additional fixation for 1 h in 1% OsO 4 , 0.8% potassium ferrocyanide, and 5 mM CaCl 2 in cacodylate buffer.EVs were ultracentrifuged at 100,000 × g for 1 h at 4°C.The resulting pellets were dehydrated with increasing concentrations of ethanol (ranging from 50% to 100%), clarified with acetonitrile, and infiltrated with araldite-Epon resin.Grids were visualized in a Jeol 100CX transmission electron microscope (JEOL, Japan) and the diameter of the EVs was calculated using ImageJ software (Bethesda, MD).Furthermore, the dimensions and concentrations of the EVs were determined using a NanoSight NS300 nanoparticle analyzer (NTA, Malvern Instruments).The instrument was equipped with a green 405 nm laser and an sCMOS camera at level 14, capturing 17.9 frames per second and a total of 1,071 frames.Viscosity was adjusted for 0.956-0.959cP (water), and measurements were conducted at 21.8°C-22.1°C.Images and data were analyzed using the NTA 3.4 software (Build 3.4.4).The dimensions of the EVs were also assessed using dynamic light scattering (DLS), with a Quasi Elastic Light Scattering in a multi-angle analyzer particle size 90Plus/ BI-MAS (Brookhaven Instruments Corp., Holtsville, NY) as previously described (49).Additionally, the values of EVs' dimensions obtained by TEM and nanoparticle tracking analysis (NTA) techniques were compared and the Pearson correlation indexes were determined, assuming a significative correlation of P < 0.05.

Metabolite, Protein, and Lipid Extraction of EVs
Three independent biological replicates of EV preparations of each A. castellanii strain (Neff and T4) were subjected to MPLEx following the protocol outlined previously (42).This consisted of adding water, chloroform, and methanol for chemical extraction and partitioning of the molecules.The EVs were then ultracentrifuged at 100,000 × g for 1 h at 4°C, followed by resuspension in the same volume of water and 5 volumes of cold chloroform-methanol solution (2:1 [vol/vol]) at −20°C.Subsequently, samples were incubated for 5 min on ice, vortexed for 1 min, and centrifuged at 12,000 rpm for 10 min at 4°C.The upper phase containing hydrophilic metabolites and the lower phase containing lipids were collected in glass autosampler vials, for metabolomic and lipidomic analyses, respectively.Proteins were recovered in the interphase and washed by the addition of 1 mL of cold methanol (−20°C), vortexed for 1 min, and centrifuged at 10,000 × g for 10 min.The supernatants were discarded, and the resulting pellets were dried in a Labconco Centrivap (Labconco, MO, USA) for 5 min.

Proteomic analysis of EVs of A. castellanii
Proteins extracted from Neff and T4 EVs were dissolved in 50 mM NH 4 HCO 3 containing 8 M urea.Protein concentration in triplicates was measured by the BCA assay (Thermo Fisher).Dithiothreitol was added to a final concentration of 5 mM to reduce the disulfide bonds at 60°C for 30 min.Then, samples were alkylated by adding iodoacetamide to a final concentration of 40 mM for 1 h at 37°C.Samples were further diluted 10-fold with 100 mM NH 4 HCO 3 previous to the addition of CaCl 2 to 1 mM final concentration.Proteins were digested with trypsin (at a 1:50 wt:wt trypsin to protein ratio) for 3 h at 37°C, followed by salts and reagents extraction in SPE columns (1 mL Discovery C18, Supelco, Bellefonte, PA) as previously described (50).Five hundred nanograms of digested peptides, dissolved in water, were loaded into a column (4 cm × 100 µm ID) packaged in-house with 5 µm C18 (Jupiter) and separated into an analytical column (70 cm × 75 μm ID) packaged with C18 particles (3 µm), connected to a nanoUPLC system (Acquity, Waters).Elution was performed with a gradient of acetonitrile/0.1% formic acid (solvent B) and water/0.1% formic acid (solvent A) over 2 h at a flow rate of 300 nL/min (51).The gradient was maintained at 1% solvent B for 15 min followed by linear increments of solvent B as follows: 19 min, 8% B; 60 min, 12% B; 155 min, 35% B; 203 min, 60% B; 210 min, 75% B; 215 min, 95% B; 220 min, 95% B. The eluents were analyzed in tandem by nanospray in an orbitrap mass spectrometer (Q-Exactive Plus, Thermo Fisher Scientific) using a scanning window of 400-2,000 m/z with a 70,000 resolution at 400 m/z.Data-dependent tandem mass spectra were acquired by high-energy collision-induced dissociation (32% normalized collision energy) for the 12 higher-intensity precursor ions with multiple loads.The dynamic exclusion function was set to exclude fragmented precursor ions for 30 s. Data acquired from liquid chroma tography coupled to tandem mass spectrometry (LC-MS/MS) were processed by the Maxquant software (v.1.6.17)(52).Peptides were identified by searching for spectra against the sequences of A. castellanii obtained from the Uniprot database (https:// www.uniprot.org/,downloaded in July 2021).Methionine oxidation and N-terminal acetylation of the protein were considered as variable modifications, in addition to carbamidomethylation of cysteine residues as fixed modifications.The tryptic digestion of both terminal peptides was considered, but two sites of lost cleavages were allowed within each peptide.The remaining parameters have been set to software default.Label-free quantification was used to compare protein abundances between different samples.

Gas chromatography analysis coupled to mass spectrometry for metabolo mic analysis in Neff and T4 EVs of A. castellanii
Hydrophilic metabolites were derivatized with methoxyamine and N-methyl-N-(trime thylsilyl)-trifluoroacetamide before being analyzed in an HP-5MS column (30 m × 0.25 mm × 0.25 μm; Agilent Technologies, Santa Clara, CA) coupled with gas chromatog raphy-mass spectrometry (GC-MS; GC 7890A/MSD 5975C, Agilent Technologies) system.The injection was performed in splitless mode at 250°C, with the temperature set at 60°C.The temperature was maintained for 1 min and continuously increased at 10°C/min to 325°C, where it was held for 5 min.The fatty acid methyl ester (SIGMA Aldrich) pattern was used as a calibrator for retention time.Recalibration, deconvolution, and data extraction were performed with a metabolite detector (55).The molecules were identified by correspondence with FiehnLib (56) and nist14 GC-MS libraries with manual curation.

Quantitative analysis of EV content from Neff and T4 strains and data integration
Comparison of the detected levels of protein orthologues, lipids, or metabolites was performed by Student's t-test, considering equal variance and two-tailed distribution, with P ≤ 0.05 values considered statistically significant.For comparative analyses, missingness was evaluated for each channel/sample, and proteins to be conserved for quantification were selected to contain at least 67% of complete values (2/3 samples) for a given condition, the overall missingness was evaluated after filtering.A heatmap depicting the proteins passing the Student's t-test (P ≤ 0.05) was plotted, and the two clusters obtained (enriched in Neff versus enriched in T4) were saved in the statistics.An enrichment analysis was performed for these two clusters by a modified Fisher's exact test (EASE test) (57-59), using the homemade Protein-Mini-On Pack age (https://github.com/GeremyClair/Protein_MiniOn).First, the statistically modulated proteins (count query) were annotated into gene ontology (GO) categories according to their molecular function, cellular localization, and biological process and also Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations and normalized to the total number of detected protein in the EVs as the universe (to calculate the % query).Second, the total number of proteins belonging to the same annotated categories in the whole genome was normalized to the universe of total genome proteins (% universe).Fold enrichment of GO or KEGG categories was calculated by dividing the (% query)/(% universe).Pathways analysis on distinct clusters was further evaluated with DAVID (60) and blast2go (https://www.blast2go.com/),and pathways were manually constructed with Vanted v2.8.3 (61).For the lipid ontology enrichment analysis, we have used the Rodin R package, available for non-R users as a web interface called Lipid Mini-On (https://omicstools.pnnl.gov/shiny/lipid-mini-on/)(62).Lipids enriched in each EV were used to perform an enrichment analysis against a universe of 185 lipid species using the Fisher exact test.For the metabolomics enrichment analysis, we used the MetaboAnalyst web interface (https://www.metaboanalyst.ca/)(63) against a universe of 86 metabolites and the Fisher exact test.In all enrichment analyses, only P ≤ 0.05 were considered statistically significant.

Sequencing and comparison of small RNAs within EVs from Neff and T4 strains of A. castellanii
A total of 100 µg of EVs were used for the RNA isolation with the miRNeasy kit (Qiagen) according to the manufacturer's instructions.The DNA cleanup step was performed in all samples using the DNase protocol (RNase-free, Qiagen).For the quantification and analysis of the integrity of RNA, we used the Qubit fluorimeter (ThermoFisher) and bioanalyzer Agilent 2100, RNA 6000 peak, and RNA small kits (Agilent Technologies).For the sRNAs and RNAs of the EVs, the libraries were built using the Kit TruSeq small RNA (Illumina) according to the manufacturer's instructions.The samples were prepared in three independent biological replicates.The RNAseq was performed on a HiSeq 2500 (Illumina, single-end 50 bp SR mid-output run).

Analysis of sequencing data
The sequences generated in fastaq format were analyzed by CLC Genomics Workbench v 20.0 (Qiagen), using the genome of A. castellanii Neff strain as a reference (NCBI RefSeq assembly GCF_000313135.1).The sequences generated were trimmed for the removal of the internal adapter (TGGAATTCTCGGGTGCCAAGG, 3' trim).The parameters for alignment were set as follows: mismatch cost (2), insertion cost (3), deletion cost (3), length fraction (0.8), and similarity fraction (0.8).The statistical analysis applied was the Differential Gene Expression method using the CLC Genomics Workbench v 20.0 (Qiagen).The RNA-seq analysis was conducted with specific parameters: the strand setting was configured for both strands, the library type set was designated as bulk, and expression was computed for genes lacking a transcript (Yes).The trimmed mean of M values was used for library size normalization and the mapping settings, the EM (expectation-maximization algorithm) estimation algorithm was employed.The differential expression analysis was based on a multi-factorial statistical approach using the negative binomial Generalized Linear Model.The false discovery rate (FDR) adjusted P-value was utilized as a multiple-testing correction.The transcripts expression was described as Transcripts per Million.The criteria for identifying differentially expressed transcripts were set at a minimum of twofold change and an FDR equal to or below 0.05.

Statistical analyses
Statistical analyses were performed using GraphPad Prism 8. Comparisons between two groups were made by the Student's t-test and P ≤ 0.05 values were considered statistically significant.Each experiment was repeated at least three times.

Comparison of growth kinetics between the Neff and T4 strains of A. castellanii
The growth kinetics of the environmental Neff and the clinical T4 strains of A. castellanii were compared up to 72 h culture.Despite the initially faster growth of the Neff strain, no statistical difference was observed for the cell density at the time of EVs harvesting (48 h), with 1.2 × 10 6 trophozoites/mL for the Neff and 9.8 × 10 5 trophozoites/mL for the T4 strain (P = 0.63, Fig. S1).

Morphological comparison of EVs from Neff and T4 strains of A. castellanii reveals distinct dimensions
TEM was utilized for morphometric analysis assessing the diameter of EVs from environmental Neff and clinical T4 strains of A. castellanii.EVs were categorized based on the size as small (≤200 nm), medium (between 200 and 400 nm), and large EVs (>400 nm), as recommended by the MISEV 2018, its updated 2023 version, and other experts in the EV field (Fig. 1A) (47,48,64,65).The histogram of size versus frequency for three independent EV preparations is shown in Fig. 1B and C for the Neff and T4 strains, respectively.A total of 410 Neff EVs were measured, with diameters ranging from 57.7 to 597 nm (mean diameter ± SD = 196.7 ± 12.0 nm).For the T4 clinical strain, a total of 594 EVs were evaluated, with diameters ranging from 27.4 to 990 nm (mean diameter ± SD = 178.2 ± 39.2 nm; Fig. 1C; P < 0.0001).Overall, a population of small EVs comprised 64.6% of Neff EVs compared to 77.3% of T4 EVs (139.52.4 nm; P = 0.11), and large EVs constituted 5.1% and 3.2% (458.1 ± 52.2 nm versus 560.9 ± 171.4 nm; P < 0.0001), respectively (Fig. 1D).

Dimensions and size distribution of Neff and T4 EVs by NTA and DLS dis played similar values to TEM
Assessment of the dimensions and concentrations of A. castellanii EVs from the Neff and T4 strains was performed using NTA.EVs isolation yielded similar concentrations as 1.1 × 10 9 ± 2.1 × 10 7 particles/mL and 1.3 × 10 9 ± 1.0 × 10 7 particles/mL, for the Neff and T4 strains, respectively.This resulted in approximately 4.5 × 10 8 ± 8.2 × 10 6 EVs/mL for the Neff and 5.1 × 10 8 ± 4.1 × 10 7 EVs/mL for the T4 strain at 48 h culture.Considering the cell concentrations measured by the growth kinetics, this renders about 388 ± 7 EVs/ trophozoite for the Neff and 519 ± 41 EVs/trophozoite for the T4 strain.For both strains, EV populations corresponded to the size found in the literature representing small EVs (30-200 nm) and medium/large EVs (>200 nm), The Neff strain EVs had a size mean of 198.0 ± 74.0 nm considering the triplicate measurements (Fig. 2A), while the T4 strain EVs had a size mean of 177.5 ± 56.6 nm (Fig. 2B).When comparing both TEM and NTA techniques for measuring the EVs dimensions, we observed a significant correlation of EVs average (Fig. 2C, P = 0.98).Additional size evaluations were performed using the DLS EVs, respectively.(F, G) Conventional principal component analysis (PCA) demonstrating grouping similarity of proteomic data in EVs replicates from Neff (green squares) and T4 (red squares) strains and (G) a comprehensive heatmap illustrating proteins with significant differences in expression (P < 0.05, downregulated in green and upregulated in red) found in EVs from Neff and T4 strains.Analysis of EVs from each strain was performed in triplicate.The statistical analysis was performed using the t-test, with P < 0.05 considered statistically significant.
technique.The Neff strain had a population of small EVs with a diameter of 45-90.8nm and a population of small/medium EVs with diameter 179.6-273.9nm, with an average diameter of 200.3 ± 56.6 nm (Fig. 2D).The T4 strain had a population of small EVs of 39.8-98.8nm and small/medium EVs of 168.5-296.5 nm, with an average diameter of 195.6 + 70.9 nm (Fig. 2E).

Assessment of protein and ergosterol levels of EVs from Neff and T4 strains of A. castellanii
Protein and sterol levels were compared between Neff and T4 strains to initially discern differences in EV content (Fig. 3A through C).Although EVs from the clinical T4 strain showed slightly lower levels of protein and sterol (Fig. 3A and B, respectively), no differences were observed regarding the protein/sterol ratio compared to the Neff strain (Fig. 3C).

Proteomic analysis reveals distinct protein cargos within EVs from Neff and T4 strains of A. castellanii
Proteomics analysis was used to compare the composition of EVs from Neff and T4 strains of A. castellanii.The results suggest potential shifts in global metabolic regula tion and phenotypic behavior.A total of 1,352 proteins were identified in combining both strains, with 1,107 of these proteins commonly found in EVs of both strains, 161 exclusively found in Neff EVs, and 84 exclusively found in T4 EVs (Fig. 3D; Table S1).The volcano plot illustrates the significance levels and expression of each protein (Fig. 3E), while the principal component analysis (PCA) of the data set reveals distinct sample grouping based on EVs from different strains (Fig. 3F).Among the proteins identified, 448 exhibited statistically significant differential expression (P < 0.05), with 69 unique proteins found in Neff EVs, 52 in T4 EVs, and 327 proteins differentially expressed in both strains.Within the latter group, 102 proteins were upregulated in Neff EVs, and 225 were upregulated in T4 EVs (Fig. 3G).
GO and KEGG mapping and enrichment analyses were conducted on the differentially expressed proteins in EVs from the Neff (171 protein) and T4 (277 proteins) strains of A. castellanii (Fig. S2), by comparing the percentage of detected proteins within each GO term to their frequency in the A. castellanii genome (Fig. 4).Both strains' EVs showed enriched proteins associated with molecular functions like ATP binding (GO:0005524), ATPase-coupled transmembrane transport activity (GO:0042626), and peroxidase activity (GO:0004601).Additionally, EVs from the Neff strain displayed proteins specifically related to serine/threonine kinase activity (GO:0004674) and phosphatidylinositol binding (GO:0035091; Fig. 4A; Fig. S2A), whereas T4 EVs had proteins associated with metal ion (GO:0046872), magnesium ion binding (GO:0000287) and actin (GO:000379), and actin filament (GO:0051015) binding (Fig. 4B).In terms of biological process, Neff EVs showed enriched proteins involved in intracellular protein transport (GO:0006886), TOR signaling (GO:0031929), and endocytic recycling (GO:0062456).The majority of proteins belonged to the integral component of the membrane (GO:0016021) cellular component (Fig. 4A; Fig. S2A).Conversely, proteins in T4 EVs were associated with biological process such as Arp2/3 complex-mediated actin nucleation (GO:0034314), response to osmotic stress (GO:0006970) and chorismite metabolic process (GO:0046417), and cellular components such cytoplasmic (GO:0005737), phagocytic vesicle membrane (GO:0030670) and clathrin-coated groups (GO:0005905, GO:0030130, and GO:0030132) (Fig. 4B; Fig. S2B).While enrichment analysis was also performed for the KEGG function for both strains, insufficient data on enriched protein groups was found for Neff EVs, with a single KEGG term displaying enrichment (path: acan00330, arginine, and proline metabolism; Fig. 4C).In contrast, several KEGG terms exhibited enrichment for proteins in the clinical T4 EVs indicating distinct participation of proteins in various biosynthetic pathways and active endocytosis (path: acan04144; Fig. 4D).
Overall, the enrichment analysis unveiled a more complex composition of T4 EVs (Table 2; Fig. S3A), with abundant glycerophospholipids such as PC, specific saturated fatty acids of 19:0 and 28:0 chains, and unsaturated 30:1 chain, PEs with a total number of chain unsaturation of 1 and TGs with a total chain carbon number of 53.On the other hand, Neff EVs were enriched with more unsaturated glycerolipids with a total number of chain carbon of 60, predominantly containing the 20:3 chain (Table 2; Fig. S3B).

Metabolite enrichment reveals distinct pathways involved in the EVs production in different A. castellanii strains
The metabolomic analysis of EVs from the Neff and T4 strains of A. castellanii identified a total of 86 metabolites (Table S4).Among these, 51 were more abundant in Neff EVs, while 35 were higher in T4 EVs.Considering the statistically significant differences in detection (P < 0.05), 11 metabolites were identified-eight predominant in Neff EVs and three in T4 EVs. Figure 7A displays a heat map illustrating the enriched metabolites.L-glutamine, ethanolamine, and glycerol were the three most prevalent species in Neff EVs, while cellobiose, maltose, and L-methionine sulfoxide were the most prevalent in T4 EVs.Enrichment analysis using the significant metabolites was performed based on the common metabolic pathway described in KEGG (https://www.genome.jp/kegg/pathway.html)(Fig. 7B and C).Phospholipid and triacylglycerols biosynthesis pathways were enriched in Neff EVs, along with the glycerolipid metabolism.In contrast, T4 EVs showed enrichment only in methionine metabolism.Fig. 7D presents a comprehensive diagram integrating detected species in EVs of both strains and statistically enriched pathways.Dihydroxyacetone phosphate and glycerol-3-phosphate, enriched in Neff EVs, are intermediates of the triacylglycerol and cardiolipin biosynthesis, glycerol phosphate shuttle and glycerolipid metabolism, along with the mitochondrial electron transport chain.Also, ethanolamine, more present in Neff EVs, is involved in the phospholipid biosynthesis, specifically of PE species.On the other hand, L-methionine sulfoxide, more abundant in T4 EVs, is an oxidation product of methionine.

Detection of RNA molecules reveals distinct transcripts in EVs of different A. castellanii strains
Most of the RNA molecules detected in the EVs ranged in size from 20 to 500 nt; however, we could also identify full-length mRNAs in the EVs.Analysis of the RNA contents of EVs via RNA-seq detected a total of 9,795 transcripts.Among these, 9,102 transcripts were present in the EVs of both strains of A. castellanii, with 339 transcripts showing differential abundance, as 180 were enriched in Neff EVs and 159 enriched in T4 EVs (Fig. 8A).Notably exclusive transcripts were detected in both EVs, with 464 found only detected in Neff EVs, and 229 in T4 EVs, and from this latter group, only five transcripts demonstrated statistical difference (Fig. 8A and B).Enrichment analysis of the pool of exclusive and differentially expressed transcripts in each EV indicates distinct transcript scenarios.In Neff EVs, transcripts involved in the biological processes of gluconeogenesis (GO:0006094) and translation (GO:0006412) were enriched, along with species typically found in the nucleosome (GO:0000786) and ribosome (GO:005840), contributing to various functions like structural constituent of ribosomes (GO:0003735), monooxygenase and oxidoreductase activity (GO:0004497 and GO:0016705, respectively) and protein heterodimerization activity (GO:0046982) (Fig. 8C).In the other hand, T4 EVs were enriched with transcripts involved in intracellular signal transduction (GO:0035556), along with molecular functions related to protein serine/threonine kinase activity (GO:0004674), ubiquitin-protein transferase activity (GO:0004842), guanyl-nucleotide exchange factor activity (GO:0005085) and microtubule binding (GO:0008017) (Fig. 8B  and D).

Data integration reveals distinct EV content from Neff and T4 strains of A. castellanii and different metabolic processes
Enrichment analysis and data integration revealed significant differences in enzyme profiles related to central carbon metabolism in EVs from different A. castellanii strains (Fig. 9).This suggests variations in metabolic processes such as glycolysis, the TCA cycle,  glycerophospholipid and glycerolipid metabolism and amino acid metabolism, which is more pronounced in the T4 strain, while the Neff strain may exhibit gluconeogenesis.Additionally, the higher levels of enzymes related to the synthesis of aminoacyl-t-RNA, nucleocytoplasmic transport, and mRNA surveillance pathway in T4 EVs, versus riboso mal biogenesis in Neff EVs indicate distinct mechanisms of transcription regulation (Fig. S4).Moreover, the elevated levels of clathrin and Arp2/3 markers in T4 EVs, along with late endosomal markers (in contrast to early endosomal markers in Neff EVs), might suggest different origins of these EV pools.The increased mTor levels in EVs of the Neff strain imply a potential autophagy process and regulation in response to nutrient starvation, while T4 EVs show higher levels of phagocytic markers (Fig. S5).The overall observations of this study are summarized in Fig. 10.

DISCUSSION
Advancements and application of new techniques have lately revealed the importance of EVs across various aspects of cell biology (66,67).These nanostructures are now recognized not only for the simple cellular cargo transportation to the extracellular environment but also for cellular-level functions, including regulation of cellular metabolism, protein expression, and even more complex processes such as quorum sensing and populational density regulation (68, 69).
Closely analyzing published articles, the vast majority of EVs secreted by parasites are small EVs, typically ranging from 25 to 100 nm in diameter.These small EVs play roles in host-parasite interactions and have direct implications for the metabolic regulation of host cells due to the presence of microRNA components (72,73,81,82).Moreover, they gain direct access to the host's intracellular environment (83).For instance, D. discoideum has been reported to secrete nanovesicles characterized by TEM as three main popula tions: nanoEVs (<50 nm), directly involved in transit from the Golgi to the extracellular environment, and two small EVs populations (between 50 and 150 nm and >150 nm) (84).
TEM morphological analysis of the EVs from Neff (57.7-597 nm) and T4 (27.4-990 nm) strains of A. castellanii under axenic cultures overlapped previously reported values for the ATCC30234 in PYG (31.9 to 467 nm) and nutritional stress (33.7 to 303.2 nm) (37,39).EVs' dimensions were also confirmed by NTA and DLS.Both techniques rely on the characteristic Brownian movement of nanoparticles in suspension.However, DLS detection is based on the scattering of light depending on particle movement and size, which is sensed by detectors placed at different angles, while the NTA particle trajectory and the scattered light upon illumination are documented by a camera strains of A castellanii.As the literature lacks information regarding specific markers to define the origin and biogenesis of EVs from A. castellanii, we followed the guidelines recommended by the MISEV 2018 (48) and its updated 2023 version (65).(85).Classically, the DLS drawback is the requirement of monodispersed suspensions; otherwise, data reliability is compromised as the measurement of particle dimensions is strongly influenced by larger particles that scatter more light (86).Then, DLS has been recommended for accurate determination of EV concentration, whereas NTA is for size determination with higher resolution (85,87).Overall, all techniques show that the Neff strain had a less polydisperse EV population, as opposed to T4 EVs, which also had a population of larger EVs.Other studies also demonstrated the presence of EVs in other A. castellanii genotypes, such as the T5 genotype showing small EVs (50.29 ± 8.49 nm and 184.6 ± 50.80 nm) at 28°C, similar to both strains used in this study (40).
To date, reports on the compositional analysis of EVs from A. castellanii are scarce, primarily focusing on proteomics.Our group described the EV protein content, comparing their composition simulating different niches A. castellanii could be found in vivo (37).Under rich nutritional conditions, A. castellanii secreted EVs containing proteins involved in signaling pathways, suggesting their involvement in cell-to-cell communica tion as previously described for D. discoideum (88,89).However, nutritional stress in the absence of protein sources caused a shift in EVs' protein contents, with most proteins related to protein and amino acid and carbohydrate metabolism, proteases, cellular stress, and oxidative metabolism, with potential implications for nutrient acquisition, colonization of distinct host niches (37), and the amoebae differentiation between trophozoite and cyst (90).
The literature reports that pathogenic strains of the T4 genotype of A. castellanii, clinically associated with keratitis and GAE, express high levels of serine proteases, whereas other genotypes express metallo and cysteine proteases (37,91).These are also secreted through EVs (36,37,41), and may pose direct implications in pathogenesis, promoting amoeba invasion digestion of extracellular matrix and mucosal protective barriers, evasion of the host's immune response, host cell death and tissue damage, ultimately granting access of the parasite to the retina, where it can cause ocular keratitis or the central nervous system, causing encephalitis (92,93).
Regarding the profile of other biomolecules and comparison among isolates within the T4 genotype, there is no report in the literature to date.This underscores the necessity for the application of multi-omic techniques such as the MPLEx and tran scriptomics, coupled with integrational analysis, for the detection of EV components.We would like to reinforce that we sought to evaluate in this study the full compo sition at multiple molecular levels, including proteins, lipids, metabolites, and small RNAs, and comprehensive characterization of the A. castellanii EVs obtained from two distinct environmental and clinical strains upon standardized 48 h cultures and starting inoculum, to infer about distinct mechanisms involved in A. castellanii pathogenesis (94).We understand that variables such as inoculum, time of culture, and medium could impact the composition of EVs, but these should be evaluated and established singularly for each strain before a multifactorial comparison could be performed.
Proteomic analysis shed light on the distinct protein cargo within EVs from the Neff and T4 strains.For both strains, we identified a total of 1,352 proteins, with 1,107 proteins commonly found in both EVs, 161 proteins exclusively expressed in Neff EVs, and 84 proteins exclusively expressed in T4 EVs.These levels reflect a 10-fold higher number of proteins than previously reported by our group (37) and Liu et al. (39), due to the use of a significantly more sensitive technique/equipment.The proteo mic findings were further substantiated by GO and KEGG mapping, with differences in the enriched proteins associated with specific biological and cellular processes, and molecular functions, suggesting potential variations in metabolic regulation and phenotypic behavior between the Neff and T4 strains.Overall, the proteins secreted and enriched in the EVs were primarily related to cellular trafficking and signaling pathways, but with distinct compositions depending on the strain.In accordance with our data, a prevalence of ATP-binding proteins was also found in the ameba D. discoideum, demonstrating their important role in amoebic survival (95).Additionally, a predomi nance of serine protease in the EVs of both but enriched in the clinical T4 strain was observed, in alignment with previous reports of their high expression in clinical strains of the T4 genotype (37,41,91).
EVs lipidomics has emerged as a novel tool to evaluate and compare alterations in the relative levels of lipid species in EVs which may vary by organism, cell type, growth conditions, and physiological state (96).Changes in the lipidome of EVs could provide information regarding EV origin and fate, trafficking, function, intracellular transport, and their stability in distinct environments, with direct implications for understanding differences in the cellular metabolism and their possible mechanisms of interaction with other cells, shedding light on their role in the pathogenesis of several diseases (97,98).Recent studies have also demonstrated that some lipids are preferentially allocated or exclusive to specific populations of EVs, serving as potential biomarkers to understand the biogenesis and packing of these structures.
A comparison of the lipid composition of EVs from the Neff and T4 strains of A. castellanii demonstrated some striking differences, and overall environmental Neff EVs displayed a prevalence of unsaturated lipid species in comparison to mostly monounsa turated lipids for the clinical T4 strain.As these lipids have significant cell membrane structural functions, we can assume that this difference in arrangement between lipids modifies membrane permeability for each strain.Neff EVs had higher levels of TG glycerolipids, with enrichment of those with a total number of 60 carbons, whereas T4 EVs displayed more DGs and DGTSAs.These glycerolipids are commonly found in lipoproteins and lipid droplets, and the higher number of TGs in the environmental Neff EVs could indicate possible lipidic reserves, as opposed to its hydrolyzed DAGs in T4 EVs (99).Glycerophospholipids such as PC are usually more enriched in the cellular membranes than in EVs of eukaryotic cells.Their enrichment in the EVs of the T4 clinical strain as compared to the Neff EVs could indicate the distinct cellular origin of EEVs between strains.Consonantly, T4 EVs were enriched in Cer (d18:1/16:0), which could suggest the promotion of microEVs secretion by this strain, as also observed in other models (100,101).Instead, Neff EVs were enriched in PS and polyunsaturated PEs, classical features of MVB-derived EVs, and exosomes that may also be involved in cell-to-cell communication and internalization of EVs by the recipient cells (102)(103)(104).However, the lack of specific markers to define the origin and biogenesis of the distinct populations of EVs, as exosomes or microEVs, for each strain used in this study, does not allow us to move forward at this stage in such characterization.
The metabolomic analysis revealed significant metabolite differences between the strains.Neff EVs displayed a similar profile to that of E. histolytica, which in a stress situation had a significant expression of glycerol and glycerol-3-phosphate, redirection of the glycolytic pathway, characteristics not observed for the clinical isolate T4.This indicates a difference in the regulation of metabolic pathways between strains (105).Elevated levels of ethanolamine in Neff EVs might also indicate a regulatory mechanism toward the synthesis of TG and PE, as part of the enriched pathways of glycerolipid metabolism, de novo triacylglycerol, and phospholipid biosynthesis.
The presence of RNAs in EVs might indicate their involvement in various genetic processes and regulation of cell physiology.Non-coding RNAs in EVs might display important regulatory functions upon delivery to the recipient cell, just as mRNA might also be translated into functional peptides once delivered, both ultimately shaping the transcriptome of the recipient cell (106,107).In Neff EVs, transcripts related to gluconeo genesis (GO:0006094) might reinforce a metabolic shift, along with transcripts related to translation (GO:0006412), nucleosome (GO:0000786), and ribosome (GO:0005840), which might explain a populational gene regulation strategy.In contrast, in T4 EVs, proteins related to signal transduction (GO:0035556) might indicate a prompt response for environmentally-induced metabolic adaptation.
Upon analyzing the outcomes of multi-omic and morphology of EVs from the two A. castellanii strains, Neff and T4, distinct regulatory mechanisms governing metabolic pathways and EV genesis become apparent (Fig. 10).This distinction was notably evident in the differential expression observed between the strains concerning the diversity of lipid classes, with the EVs of the clinical strain T4 manifesting six distinct lipid classes compared to the three in the environmental Neff strain.Moreover, variations in unsaturation levels, notably observed in phosphatidylethanolamines, further delineated the unique lipid profiles characterizing each strain.This lipidomic variance is corrobo rated by the functional differences of proteins identified in the strains, aligning with specific metabolic processes.The Neff strain exhibits a higher concentration of proteins associated with autophagy and the regulation of nutrient deficiencies, highlighting a robust internal regulatory machinery.Conversely, the T4 strain shows enrichment of proteins involved in phagocytic processes, reflecting a pronounced focus on interactions with and adaptation to the external environment.This contrast is similarly observed in RNA analysis, demonstrating a heightened internal regulatory activity in Neff and greater adaptability to diverse metabolic pathways in T4.
In this broader context, we observed in general that the EVs of the Neff and T4 strains of A. castellanii have different compositions due to the distinct behavior of their generating cells in response to their respective environment.The environmental strain displayed a rigorous regulation of mechanisms involved in the internal machinery of the cell, while for the clinical strain T4 there was a greater presence of markers aimed at interaction with the external environment, as evidenced by phagocytosis markers.This divergence in behavior offers valuable insights into the different damage profiles the two strains cause during infection in a host.
Thus, our results reinforce the need for further multi-omic studies targeting EVs as critical pathogenic mechanisms thanks to their ability to carry molecules with high virulent potential to the external milieu.In addition, comparative analysis of EV composition across various strains enables the exploration of secretion pathways as potential pharmacological targets for the treatment of A. castellanii infections.With the application of increasingly integrative techniques, such as MPLEx used in the current work, we have been able to further advance the knowledge of the plasticity of metabolic pathways linked to the virulence of diverse pathogens, as is the case of A. castellanii.

ADDITIONAL FILES
The following material is available online.

FIG 1
FIG 1 TEM images of EVs from the environmental Neff (ATCC30010) and clinical T4 (ATCC50370) strains of A. castellanii demonstrating variations in population sizes between strains.(A) EVs including exosomes and microvesicles from Neff and T4 strains.(B and C) Histograms depicting the frequency of EVs diameter by TEM (triplicates) for (B) Neff and (C) T4 strains, upon TEM and measurement in ImageJ.(D) Comparison of the mean EV diameter for Neff and T4 strains (each triplicate).(E) Scatter plot displaying EVs diameter of Neff and T4strains by TEM.Images captured at 46,000× magnification.

FIG 2
FIG 2 Frequency distributions of EVs diameters for the Neff and T4 strains of A castellanii, as determined by NTA and DLS.(A and B) Histograms presenting the frequency of EVs diameter using NTA for the (A) Neff and (B) T4 strains.(C) Mean diameter comparison of EVs from Neff and T4 strains obtained via TEM and NTA.(D and E) Frequency plots of EVs diameters of (D) Neff and (E) T4 strains using DLS.The red dashed lines indicate the mean diameter obtained for the EVs from each strain.

FIG 3
FIG 3 Biochemical and proteomic characterizations of EVs from Neff and T4 strains of A. castellanii.(A) The concentration of proteins found in the EVs of Neff and T4 strains.(B) Concentration of sterols found in the EVs of Neff and T4 strains.(C) Protein/sterol concentration ratio found in the EVs of the Neff and T4 strains.(D) Venn diagram illustrating the proteomic analysis of EVs of environmental (Neff) and clinical (T4) strains showing total, strain-specific, and shared proteins between the strains (and those with statistically significant differential expression, P < 0.05).(E) Volcano plot showing the magnitude of change (fold change T4/Neff) versus statistical significance (adjusted P value); dark green and red dots denote the differentially expressed proteins, more abundant in Neff and T4

FIG 4
FIG 4 Enrichment analysis of proteins differentially regulated in EVs from Neff and T4 strains of A. castellanii.(A and B) Gene ontology enrichment analysis of the biological process (green), cellular component (pink), and molecular function (blue) for proteins more abundant (in EVs of the (A) Neff and (B) T4 strains of A. castellanii.Legends on the right panel (upper lane) for both (A and B) indicate the symbol size (fold change) (C and D) KEGG pathway enrichment analysis for proteins more abundant in EVs of the (C) Neff and (D) T4 strains of A. castellanii.

FIG 5
FIG 5 Analysis of total lipid composition obtained under the positive ionization mode of EVs from Neff and T4 strains of A. castellanii.(A) Number of detected lipids in positive ionization mode for EVs from Neff and T4 strains, with the statistically significant increased species (P < 0.05) number for each strain displayed in the green and red areas, respectively.(B) Heatmap of significant lipids (P < 0.05) grouped by lipid classes found in the positive mode analysis of EV from Neff and T4 strains of A. castellanii.Values were calculated after normalization of intensities by the mean values obtained for the environmental isolate Neff {log2[(sample)/(Neff average)]}.(C) Scheme denoting the fatty acid chains and unsaturation profiles found among the significant lipid classes obtained in the positive mode analysis of EVs from the Neff (green rectangles) and T4 strains (red rectangles).TG, triacylglycerol; DG, diacylglycerol; DGTSA, diacylglycerol-trimethylhomoserine; PC, phosphatidylcholine.

FIG 6
FIG 6 Analysis of total lipids composition obtained under the negative ionization mode of EVs from Neff and T4 strains of A. castellanii.(A) Number of detected lipids in negative ionization mode for EVs from Neff and T4 strains, with the statistically significant increased species (P < 0.05) number for each strain displayed in the green and red areas, respectively.(B) Heatmap of significant lipids (P < 0.05) grouped by lipid classes found in the negative mode analysis of EV from Neff and T4 strains of A. castellanii.Values were calculated after normalization of the intensities by the mean values obtained for the environmental isolate Neff {log2[(sample)/(Neff average)]}.(C) Scheme denoting the fatty acid chains and unsaturation profiles found among the significant lipid classes obtained in the negative mode analysis of EVs from the Neff (green rectangles) and T4 strains (red rectangles).PS, phosphatidylserine; PE, phosphatidylethanolamine; Cer, ceramide; PA, phosphatidic acid.

FIG 7
FIG 7 Analysis of metabolites for which significant statistical significances (P < 0.05) in EVs from the Neff and T4 strains of A. castellanii.(A) Heatmap of differentially abundant (P < 0.05) metabolites found in EVs from Neff and T4 strains.Values were calculated after normalization of the intensities by the mean values obtained for the environmental isolate Neff {log2[(sample)/(Neff average)]}.(B and C) KEGG pathway enrichment analysis for metabolites more abundant in EVs of the (B) Neff and (C) T4 strains of A. castellanii.(D) Diagram demonstrating the integration of lipid metabolic pathways of differentially abundant metabolites between the EVs from Neff (black bars) and T4 (gray bars) strains of A. castellanii.Green squares: metabolites more abundant in Neff EVs; red squares: metabolites more abundant in T4 EVs.

FIG 8
FIG 8 Analysis of the microRNA contents of EVs from Neff and T4 strains of A. castellanii using RNA-seq.(A) Venn diagram demonstrating the differential numbers of small and mRNAs transcripts detected in EVs from Neff and T4 strains of A. castellanii.(B) Heat map of the differentially regulated transcript (344), with 180 more abundant in Neff EVs, 159 more abundant in T4 EVs, and five exclusive transcripts found in T4 EVs.(C and D) Gene ontology enrichment analysis of the biological process (green), cellular component (pink), and molecular function (blue) for transcripts more abundant in EVs of the (C) Neff environmental and (D) T4 clinical strains of A. castellanii.

FIG 9
FIG 9 Diagram illustrating the levels of differentially regulated proteins (square graphs) and abundant metabolites (circles) participating in carbon metabolism and accessory metabolic pathways in the EVs of environmental Neff (black bars) and clinical T4 (gray bars) strains of A. castellanii.Green squares: proteins more abundant in Neff EVs; red squares: proteins more abundant in T4 EVs; yellow circles: metabolites more abundant in Neff EVs, and pink circles: metabolites more abundant in T4 EVs.

FIG 10
FIG 10 Diagram summarizing the results obtained throughout all techniques applied to compare the EVs composition of the environmental Neff and clinical T4

TABLE 2
Lipid Ontology enrichment analysis of differentially regulated lipids in EVs from Neff and T4 strains of A. castellanii using the Lipid Mini-On.