Characterization of Extracellular Vesicles Produced by Aspergillus fumigatus Protoplasts

Fungal cells use extracellular vesicles (EVs) to export biologically active molecules to the extracellular space. In this study, we used protoplasts of Aspergillus fumigatus, a major fungal pathogen, as a model to evaluate the role of EV production in cell wall biogenesis. Our results demonstrated that wall-less A. fumigatus exports plasma membrane-derived EVs containing a complex combination of proteins and glycans. Our report is the first to characterize fungal EVs in the absence of a cell wall. Our results suggest that protoplasts represent a promising model for functional studies of fungal vesicles.

which allows this organism to successfully establish disease in different hosts (1)(2)(3)(4). Aspergillosis begins with inhalation of asexual conidia followed by fungal morphological transition in the absence of a proper immunological response (1).
A major difficulty in directly addressing the physiological roles of EVs is the lack of understanding of their intracellular biogenesis and of the mechanisms underlying cell wall crossing. An original approach for studying the generation of EVs would be to use protoplasts and look at their active release in the absence of a cell wall. In fact, protoplasts might represent a promising model for the study of EVs, as already suggested by Gibson and Peberdy 50 years ago, who chose the name "subprotoplasts" for the vesicle-like particles budding from the plasma membrane of A. nidulans cells (23).
Our primary goal in the present work was to search for EVs produced by protoplasts of A. fumigatus germinating conidia. Our results revealed the presence of typical EVs in protoplast supernatants incubated under different conditions. EV cargo was directly influenced by the experimental conditions under which A. fumigatus was incubated. Our report provides experimental evidence that fungal EVs are produced not only by the mycelial morphological stage of A. fumigatus, but also by protoplasts of germinating conidia, especially during cell wall regeneration. To our knowledge, this is the first demonstration that protoplasts may represent a useful model to analyze the production and role of EVs in the absence of any cell wall in an experimental setting similar to the ones which allowed the study of exosomes produced by mammalian cells (24).

RESULTS
Observation of outer particles resembling EVs in protoplasts of A. fumigatus germinating conidia. To experimentally overcome the difficulties of detecting EVs in situ due to the presence of a thick cell wall, we adopted an experimental model using protoplasts. Fungal cells lacking cell walls were obtained by enzymatic digestion with a lysing enzyme from Trichoderma harzianum (Glucanex), which hydrolyzes cell wall components for protoplast preparation (25). These cells can reconstruct their walls in osmotically stabilized media containing 0.6 M KCl, glucose, nitrogen, and salts (26)(27)(28)(29)(30). Incubation of protoplasts in 0.6 M KCl only, on the other hand, impairs cell wall synthesis (A. Beauvais and T. Fontaine, unpublished data). In this context, we first compared the morphological aspects of freshly obtained A. fumigatus protoplasts with cell wall-regenerating cells by scanning electron microscopy (SEM). Under both experimental conditions, we observed ϳ50-nm-diameter extracellular structures in an apparent association with the fungal surface (Fig. 1). When the protoplasts were incubated under conditions of cell wall synthesis, the outer particles were more numerous, and a fibril-like network was observed.
To analyze these vesicle-like particles in intact protoplasts, we checked the membrane organization of freshly prepared protoplasts, protoplasts undergoing cell wall synthesis for 2 h (medium with the osmotic stabilizer KCl and nutrients), and protoplasts incubated for 2 h under the starvation condition (in KCl only). Incubation outcomes were monitored by staining the protoplasts with an anti-glucan antibody and observation by fluorescence microscopy. Membranes were stained with the lipophilic dye DiI. A. fumigatus protoplasts manifested the complex membrane distribution that is typically observed in most eukaryotic cells ( Fig. 2A). As expected, glucan was detected at the background levels in freshly prepared protoplasts and in protoplasts incubated in KCl alone. Under conditions of cell wall synthesis, surface glucan was unequivocally detected. A detailed analysis of the relationship between membrane staining and glucan detection in these cells revealed several regions of membrane projection to the outer space (Fig. 2B). Under nonregenerating conditions, no glucan was detected (Fig. 2C). During induction of cell wall synthesis, the projected regions were closely associated with glucan detection, and, in fact, the polysaccharide was apparently surrounded by the membranous compartments (Fig. 2D). The occurrence of membrane projections in protoplasts incubated under nonregenerating or cell wall synthesis conditions was confirmed by superresolution SEM (Fig. 2E and 2F, respectively).
Protoplasts of A. fumigatus germinating conidia release EVs. Fungal preparations used for both qualitative and quantitative EV analyses were adjusted to 10 8 protoplasts/ml, and cell viability was in the 87% to 93% range throughout all of the experiments. Supernatants obtained from protoplasts incubated under regenerating and nonregenerating conditions were fractionated by ultracentrifugation, and the resulting pellets were analyzed by transmission electron microscopy (TEM). Membranous structures with a typical bilayer and with the morphological aspects and dimensions previously observed for fungal EVs were isolated from protoplast supernatants (Fig. 3A, B, E, and F). Regenerating protoplasts produced EVs that were apparently associated with fibrillar material (Fig. 3G and H; see also Fig. S1 in the supplemental material). Most of the vesicles were in the 200-nm-diameter range, and this visual perception was confirmed by nanoparticle tracking analysis (NTA), which revealed a major peak of vesicle detection in the 150-nm-diameter range ( Fig. 3I and J). EVs obtained from control or regenerating protoplasts had very similar diameter distribution profiles.
EVs are more abundantly detected in the supernatants of regenerating protoplasts. We quantified EV production in the experimental systems explored in our study by different approaches. First, independent replicates were submitted to quantitative NTA. This analysis revealed that the EV/cell ratios were at the background levels for the samples from freshly prepared protoplasts (Fig. 4A). Even though some EVs were released over time in the supernatants of control protoplasts, their number was drastically increased in the supernatant of regenerating protoplasts. The NTA data agreed with the quantification of the sterols in the EV-containing supernatants (Fig. 4B). This increase in the sterol levels was not associated with an enhancement in the total amount of the protoplast sterols occurring during cell wall regeneration (data not shown).
Glycan components of A. fumigatus EVs. The molecular composition of the EVs released by the protoplasts was investigated. Since galactosaminogalactan (GAG) is a marker of polysaccharide secretion by the A. fumigatus mycelium (4), we first showed the presence of this polysaccharide during protoplast regeneration (Fig. 5A). We then proved that GAG was present in EVs isolated from protoplasts incubated under the conditions of cell wall regeneration and was absent in control vesicles (Fig. 5B). These results agreed with the compositional analysis of the carbohydrate units of EVs. Glucosyl (Glc), mannosyl (Man), and galactosyl (Gal) units were detected in EV preparations obtained from both cell wall-regenerating and nonregenerating protoplasts (Fig. 5C). N-Acetyl-galactosaminyl (GalNAc) residues, which are markers of GAG, were found only in regenerating protoplasts. In addition, Glc levels were significantly increased in EV samples from these cells compared to nonregenerating protoplasts (P ϭ 0.004). N-Acetylglucosamine (GlcNAc) residues were absent in all samples.
Proteomic analysis of EVs. Proteomic analysis revealed only 142 proteins in EVs produced by fresh protoplasts, contrasting with the detection of 2,056 proteins in  vesicles from regenerating protoplasts (see Table S1 in the supplemental material). All 142 of the EV proteins detected in fresh protoplasts were found in samples obtained under cell wall regeneration conditions. Although EVs were more abundantly detected in the regenerating protoplasts, the qualitative protein composition of the nonregenerating protoplast was similar (Table S1). The predicted GO classification of all EV-related proteins identified numerous terms ( Fig. 6). As previously described for several fungal EVs (11,14,22,31), the shared GO terms corresponded to proteins involved in a wide range of processes of fungal physiology. Most of the biological processes (680 GO terms) were common to both the regenerating and nonregenerating conditions. A minor fraction of biological processes (50 GO terms) were specifically found in the cell wall-regenerating system. We used the UniProt (https://www.uniprot.org/) and AspGd (http://www.aspgd.org/) databases to specifically analyze proteins related to cell wall assembly under each of the sets of experimental conditions used in this study (Table 1). We detected several proteins related to (i) cell wall synthases, including the ␤1,3 glucan synthase Fks1 and its GTPase activator Rho1 (32), ␣1,3 glucan synthases Ags (33), chitin synthases (34), Ktr mannosyltransferases involved in the synthesis of galactomannan (35), and enzymes belonging to the GAG biosynthetic pathway (Ugm1, Gt4c) (36); (ii) cell wall remodeling enzymes, including Gel and Bgt glucosyltransferases involved in the elongation and branching of the ␤1,3 glucan (37); and (iii) some enzymes involved in mannosylation, including multienzyme complexes mannan polymerase I and II involved in the synthesis of one of the two conidial mannans (Mnn proteins, Van1, Och1) (38) and Pmt O-mannosyltransferases involved in the 0-mannosylation of cell wall remodeling enzymes (39). Other cell wall-related proteins, including Mp1, PhiA, and MidA; glycosylphosphatidylinositol (GPI)-anchored proteins, including Ecm33; mannosyltransferases (Alg2, Och1and MnnII); and the putative glycan biosynthesis protein Pig1, were also   detected. Of note, the proteins identified under nonregenerating conditions likely represent underestimations, as a consequence of the lower production of EVs under those conditions.

DISCUSSION
Yeast forms of different fungal species produce extracellular membrane structures classified as EVs (40). More recently, it was demonstrated that filamentous forms of fungi also produce EVs (18)(19)(20)22). Although the functional impact of these findings is still not clear, they confirm that EVs are released by fungi in different morphological stages as part of distinct physiological events. In most eukaryotes, EVs are released at the plasma membrane level. However, in fungi and plants, the cell wall is usually the outermost cell layer, increasing the complexity of understanding the physiological function of eukaryotic EVs. Our study has shown that living wall-less stages such as protoplasts are useful as models to analyze the role of fungal EVs in cell wall biogenesis. Of note, protoplasts of other species, including C. albicans, Schizosaccharomyces pombe, and Neurospora crassa, also produced extracellular particles resembling EVs (23,27,41), as concluded from previous microscopic analyses of fungal cells.
Our SEM analysis of protoplast forms of A. fumigatus germinating conidia revealed the presence of particles with general properties compatible with those of EVs being released from the plasma membrane, including morphology, dimensions, and bilayered membranes (42). These vesicular particles were morphologically similar to those observed in association with the cell wall of C. neoformans (43). This result and the observation of vesicles emerging from the A. nidulans surface (23) suggested that A. fumigatus protoplasts are efficient producers of EVs.
EVs released by regenerating protoplasts showed fibril-like material attached to the lipid surface, as revealed by TEM of isolated EVs. Even though the EVs were produced in the highest number in regenerating protoplasts, it is noteworthy that EV release was not uniquely associated with cell wall biosynthesis since the nonregenerating protoplasts also produced EVs. This observation suggests that EV release is not exclusively related to the synthesis of the cell wall and that the release of EVs in nongrowing cells may be also a response of the fungus to extracellular stress such high osmotic pressure or lack of nutrients. It is noteworthy that EVs from cell wall-regenerating cells and nonregenerating protoplasts differed in relative concentrations. Considering that these experimental systems correspond to conditions of nutrient abundancy and starvation, respectively, the quantitative differences could simply correspond to a more efficient metabolic response of nonstarved cells. Indeed, the impact of the nutritional availability on the production of microbial EVs has been reported before. In Mycobacterium tuberculosis, EV production was found to increase in response to iron restriction (44). Alternatively, the increased number of EVs in our protoplast model might indicate an association between vesicle production and cell wall synthesis. That supposition remains to be experimentally proved, but the finding that S. cerevisiae strains with deletions in cell wall biosynthesis genes produced more EVs than parental cells (45) argues against this hypothesis but favors the hypothesis that stress increases production of EVs.
Our carbohydrate analysis revealed the presence of Man and Gal, an increased amount of Glc, and the presence of GalNAc in EVs obtained from regenerating protoplasts. Glc is a marker of ␤1,3 or ␣1,3 glucan. Man and GaI are markers of galactomannan. Our current results suggest that the EV population produced by A. fumigatus includes plasma membrane-derived vesicles, as consistently described for the mammalian EVs denominated microvesicles (24). Therefore, we speculate that ␤1,3 or ␣1,3 glucan is incorporated in the EVs during their formation, as concluded from previous demonstrations that cell wall polysaccharides are synthesized on the internal side of the plasma membrane level and extruded in the cell wall at the C-terminal pore-like part of the respective enzymes (46,47). Galactomannan is assembled in the Golgi apparatus and secreted to the plasma membrane before being cross-linked to ␤1,3 glucan, supposedly by extracellular transglycosidases (35). Therefore, the possible presence of glucans and galactomannan in EVs may be a consequence of their original association with the plasma membrane. GAG, which is a virulence-associated component of the A. fumigatus extracellular matrix (48,49), is localized on the surface of the cell wall, where it acts as a component of the fungal extracellular matrix (50). The detection of GalNAc residues only in EVs obtained from regenerating protoplasts suggests that GAG is transported by vesicles through the cell wall to be deposited on the cell surface. Alternatively, GAG could be loosely associated with EVs, considering its "sticky" nature due to its great ability to form unspecific hydrogen bonds. Similar observations were described in C. neoformans, in which EVs were found to contain the extracellular polysaccharide glucuronoxylomannan (9). It is still unknown whether the EVs characterized in our protoplast model represent the vesicular structures produced by intact A. fumigatus. In our study, we identified a higher diversity of EV proteins than of A. fumigatus mycelial vesicles (22), but the hypotheses explaining these differences are numerous. First, it is important to highlight that the strains used in these studies were distinct, which impairs an accurate comparison. Second, the proteomic analyses in these independent studies were performed, as usual in the literature, under very different technical conditions. Finally, all studies on fungal EVs produced so far had used distinct conditions for fungal growth that do not correlate with our model of nutritional abundancy (cell wall synthesis) or starvation (nonregenerating conditions). Compositional comparisons between this study and others are, therefore, of very limited applicability. Nevertheless, 32 of the 60 proteins described for the mycelial A. fumigatus EVs were also found in our study.
Our current results reinforce the idea that biogenesis of fungal EVs includes vesicle formation at the plasma membrane level, as demonstrated for other eukaryotes (24). However, the presence of intracellular sites of vesicle biogenesis and their relationship with the synthesis of the cell wall cannot be ruled out. For instance, Neurospora crassa chitin synthases 1, 3, and 6 were previously shown to be distributed into cytoplasmic vesicular compartments likely corresponding to chitosomes (51). In Zymoseptoria tritici, chitin and ␤(1,3)-glucan synthases were previously found to be coexported to the cell surface within the same vesicle (52). In plant cells, the glucan synthase-like protein NaGSL1 was detected in both intracellular vesicles and the plasma membrane, the latter location being associated with cell wall synthesis (53). In the Aspergillus model, protoplasts are produced from the germ tube tips, so they contained all the cell wall synthesis and remodeling machinery normally present in the plasma membranes of fungal apexes (36,54,55). Consequently, it is not surprising that protoplast EVs contained many of these cell wall enzymes. Some of these proteins, including Fks1, CsmB and Chs, Gel4, Pmt4, Ktr4, Ktr7, and Gt4C, are essential for the synthesis of the major cell wall polysaccharides ␤1,3 and ␣1,3 glucans, chitin, galactomannan, and GAG and for branching/elongation of the ␤1,3 glucan in A. fumigatus (56)(57)(58). The absence of GlcNAc (N-acetylglucosamine residues) may suggest that chitin-related molecules are absent in the EVs. However, we cannot rule out the possibility that a slower kinetics of chitin synthesis affected our experimental model. For instance, preliminary experiments of polysaccharide immunolabeling of regenerating protoplasts showed that ␤1,3 glucan was the first polysaccharide detected on the surface of the protoplasts, following by ␣1,3 glucan. Chitin was the last polysaccharide to be detected (A. Beauvais, unpublished results).
Proteins that are not predicted to be in the extracellular milieu were abundantly detected in EVs from A. fumigatus protoplasts. This observation agrees with numerous reports on the protein composition of fungal EVs (8,11,14,31) and with the fact that the biogenesis of these membranous structures has been linked to the cytoplasm and the plasma membrane (59). Additionally, it was previously suggested that the cell wall is a storage site for many fungal proteins, including glucanases, PhiA, Ecm33, Gel1, and Gel4 (60-63). Our current results contribute to a better understanding of the properties of fungal EVs. To our knowledge, this is the first characterization of EVs in protoplasts obtained from germinating conidia. Our results suggest that these cellular forms represent a promising model to explore novel roles of fungal EVs in many fungal species. In the A. fumigatus model, phagocytic cells stimulated with EVs increased their ability to produce inflammatory mediators and to promote fungal clearance (22). Similarly, A. flavus EVs affected the interaction between the fungus and host immune cells (64). These results support the relevance of the use of protoplastic fungal EVs to promote better understanding of their role in both the physiology and immunopathogenesis of A. fumigatus.

MATERIALS AND METHODS
Growth conditions and preparation of A. fumigatus protoplasts. The A. fumigatus reference strain used in this study was CEA17ΔakuB KU80 (ku80), which is deficient in nonhomologous end joining (65). The CEA17ΔakuB KU80 strain was conserved on 2% (wt/vol) malt agar slants. Five-day-old conidia were recovered from the slants by vortex mixing performed with a 0.05% (vol/vol) aqueous Tween 20 solution and filtered through a 40-m-pore-size cell strainer. For protoplast preparation, 7 ϫ 10 9 conidia suspended in 0.05% Tween were centrifuged at 3,000 ϫ g for 10 min. The supernatant was discarded, and the cells were suspended in 10 ml of sterile water. The conidia were inoculated in 600 ml of germination medium (1% yeast extract, 3% glucose, and 0.6 M mannitol) and then incubated with shaking for 14 h at 30°C. Germinated conidia were harvested and separated from dormant conidia by filtering the sample through a sterile Miracloth-lined funnel and then washed with 200 ml of sterile osmotic medium (OM; 1.2 M MgSO 4 · 7H 2 O, 0.09 M K 2 HPO 4 , and 0.01 M KH 2 PO 4 , pH 5.8) and subjected to 2-fold dilution. After the washing step, the germinated conidia were suspended in 20 ml OM containing Glucanex (Novo Nordisk Ferment Ltd., catalog number CH4243) at 30 mg/ml. The cells were gently homogenized, and the suspension was adjusted to a final volume of 250 ml with a filtered-sterile Glucanex solution for hydrolysis of the cell wall.
Cell wall digestion was performed in 2-liter Erlenmeyer flasks for 2 h at 37°C, with gentle shaking (60 rpm), until sufficient protoplasts were released, as assessed microscopically. For protoplast recovery, the cells were harvested by filtration in a sterile glass Buchner funnel (porosity 2). After filtration, 2 volumes of sterile 0.3 M KCl were added to 1 volume of the protoplast suspension in Glucanex. The mixture was centrifuged at 5,000 ϫ g (20 min, 25°C). The pellet was suspended in sterile 0.6 M KCl and washed twice (3,000 ϫ g, 10 min per wash, 25°C, with minimal break) to eliminate the remaining Glucanex. The final pellet of fresh protoplasts was divided into two equal parts. Each part was suspended in 10 ml of sterile 0.6 M KCl and further incubated under nonregenerating or cell wall synthesis conditions. The viability of the protoplasts was assessed by the use of trypan blue dye exclusion at different time points, from 0 to 2 h.
For cell wall regeneration, the protoplasts were incubated in 400 ml of a minimal medium (MM) supplemented with 0.6 M KCl for 2 h at 37°C, with shaking (120 rpm). The MM was prepared as previously described (66,67) with some modifications and contained 1% (wt/vol) glucose, 20 mM glutamine, 0.052% KCl, 0.052% MgSO 4 · 7H 2 O, 0.152% KH 2 PO 4 , and 1 ml trace element solution (pH 6.5). Alternatively, protoplasts were incubated for 2 h in a nonregenerating solution of 0.6 M KCl or were immediately processed for the analyses described below.
Microscopic analysis of protoplasts. Freshly purified protoplasts, as well as protoplasts obtained under cell wall-regenerating or nonregenerating conditions, were fixed with 2% formaldehyde-0.6 M KCl and stored at 4°C. For fluorescence or superresolution microscopy, the cells were washed twice with phosphate-buffered saline (PBS) and blocked with superblock blocking buffer (Thermo Scientific, catalog number 37515) mixed in PBS for 1 h at 37°C. Surface ␣-(1,3)-glucan of protoplasts was labeled with MOPC 104E monoclonal antibody (Sigma-Aldrich, catalog number M5909) (2 g/ml; 1 h at 37°C) (68). Alternatively, the cells were stained with an anti-GAG monoclonal antibody (20 g/ml; 1 h at 37°C) produced in the Latgé laboratory as previously reported (48). After two washes in PBS, each preparation was incubated with the appropriate secondary antibodies (anti-mouse IgM Alexa Fluor 488 for glucan staining; anti-mouse IgG Alexa Fluor 488 for GAG; both diluted 1:100 in blocking buffer). After incubation for 1 h at 37°C, the cells were washed three times with PBS. The protoplast membranes were finally stained with Vybrant DiI cell-labeling solution (Molecular Probes, catalog number V22885) at 5 M (30 min, 37°C) and washed one final time with PBS. The cells were placed on glass slides covered with ProLong Gold antifade reagent. The cells were microscopically observed under a Zeiss Axioplan 2 fluorescence microscope or a Zeiss Elyra PS.1 superresolution microscope using structural illumination mode. Images were obtained with ZEN 2.1 software.
Superresolution scanning electron microscopy (SEM) was performed as previously described (69). Briefly, 10 6 protoplasts were fixed on sterile glass coverslips (previously coated with poly-L-lysine) overnight at 4°C on 24-well plates. The samples were dehydrated in a graded ethanol series (30%, 50%, and 70% for 5 min and 95% and 100% for 10 min), subjected to critical point drying in CO 2 , mounted on stubs, and coated with carbon. Observation of the protoplast cell surface was performed with an Auriga 40 field emission scanning electron microscope (FE-SEM) microscope (Zeiss, Germany).
Protoplasts were also were analyzed with a JEOL JSM-6700F apparatus, which is an ultra-highresolution field emission scanning electron microscope equipped with a cold-field-emission gun and a strongly excited conical lens. The secondary-electron image resolution settings were 1 nm at 15 kV and 2.2 nm at 1 kV. Pieces of culture were frozen using a Gatan Alto 2500 cryo-stage and cryo-preparation chamber. The preparation conditions were as described previously by Paris et al. (70).
EV isolation and physical-chemical analysis. Isolation of EVs from protoplast supernatants was performed as previously described for yeast cells (40), with minor modifications. Briefly, after each incubation period, the supernatants were separated from the protoplast cells by centrifugation at 3,000 ϫ g (15 min, 25°C, with no brake) and sequentially passed through filters with 5-m, 1.2-m, and 0.45-m pore sizes. The pellets containing protoplast cells were stored at Ϫ20°C for sterol quantification. After filtration, the supernatants were concentrated in an Amicon ultrafiltration system (cutoff, 100 kDa) and again centrifuged at 10,000 ϫ g and 4°C for 15 min to eliminate possible cellular debris. The concentrated supernatants were finally ultracentrifuged at 100,000 ϫ g (4°C, 1 h). The resulting pellets containing EVs were washed twice with filtered PBS (0.22-m pore size) under the same ultracentrifugation conditions and finally suspended in 300 l of filtered PBS (0.22-m pore size). The EV suspensions were stored at Ϫ80°C for further experiments.
For nanoparticle tracking analysis (NTA) and GAG serological detection, the EV suspensions were first submitted to immunoprecipitation for the removal of nonvesicular polysaccharides. In this assay, 50 l of the EV suspension was added to the wells of a 96-well enzyme-linked immunosorbent assay (ELISA) plate, previously coated with a mixture of antibodies against ␣-glucan (J558) (71), ␤-glucan [Dectin 1 human IgG Fc chimeric ␤᎑(1,3)᎑glucan receptor; a kind gift of G. Brown, University of Aberdeen, United Kingdom], and GAG (1 g/ml, 1 h, 37°C) (48) and blocked with PBS containing 1% bovine serum albumin (BSA). Unbound fractions were collected, and the resulting EVs were stored at Ϫ80°C for further experiments.
Nanoparticle tracking analysis (NTA) was performed to determine the EV diameter and concentration. NTA of protoplast EVs was performed on an LM10 nanoparticle analysis system coupled with a 488-nm-wavelength laser and equipped with a scientific complementary metal oxide semiconductor (sCMOS) camera and a syringe pump (Malvern Panalytical, Malvern, United Kingdom), as recently described for C. gattii EVs (72). The samples were subjected to 25-fold dilution in filtered PBS and measured within the optimal dilution range of 7.6 ϫ 10 7 to 6.8 ϫ 10 8 particles/ml. The data were acquired and analyzed using NTA 3.0 software (Malvern Panalytical). NTA values were used to calculate ratios of EVs to cells by dividing the number of particles detected in the equipment by the number of cells in the original individual vesicle samples. For these analyses, two independent biological replicates were prepared and each sample was submitted to at least three reads, generating six measurements for each sample. These values were adjusted according to the original sample dilution (25-fold). The quantification of sterol in EV preparations was performed with an Amplex red cholesterol assay kit (14,73,74).
Transmission electron microscopy (TEM) of EVs. For negative-staining TEM, the EV pellets were fixed with 2% glutaraldehyde-2% paraformaldehyde-0.1 M sodium cacodylate buffer at room temperature for 2 h and then postfixed overnight at 4°C with 1% glutaraldehyde-4% paraformaldehyde-PBS. Copper carbon-coated grids (Cu-CF300; EMS), previously negatively charged by the use of an Elmo system (1 min, 15 pascals, 2 mA, 80 V), were put in contact with 15 l of each sample for 10 min and washed three times with Milli-Q water drops (2 min each time), stained with uranyl acetate 2%, dried, and observed with a Tecnai Spirit microscope operating at 120 kV and equipped with an Eagle 4,000-pixelby-4,000-pixel camera.
EVs were alternatively fixed with 2% formaldehyde-2% glutaraldehyde-cacodylate buffer (0.1 M, pH 7.4). The samples were washed through four changes of cacodylate buffer (30 min each) and pelleted in 1% agarose (JT Baker Chemical Co., Phillipsburg, NJ). They were transferred to 1% osmium tetroxide in cacodylate buffer (0.1 M, pH 7.4) and incubated at 4°C for 1 h followed by washing in cacodylate buffer and distilled water performed for a total of 30 min. The samples were then stained with 0.5% aqueous uranyl acetate, dehydrated, slowly infiltrated with epoxy, and embedded. After resin polymerization, the blocks were sectioned on a Leica ultramicrotome and subjected to poststaining for 10 min in 2% uranyl acetate-50% ethanol and for 5 min in lead citrate. Ultrathin sections (70 nm thick) were collected on Formvar-coated copper slot grids and subjected to poststaining for 10 min in 2% uranyl acetate-50% ethanol and for 5 min in lead citrate. Sections were then examined on a JEOL 1200EX microscope (JEOL Ltd., Tokyo, Japan) equipped with a SIA L3C charge-coupled-device (CCD) camera (SIA Inc., Duluth, GA).
Monosaccharide composition and serological detection of GAG in EVs. EV ultracentrifugation pellets were suspended in water for monosaccharide analysis. The presence of monosaccharides in the EVs was determined by gas chromatography after hydrolysis, reduction, and paracetylation of the vesicle components using meso-inositol as an internal standard (75). Serological estimation of vesicular GAG levels was performed as described before by our group for other fungal polysaccharides (74). Briefly, EV suspensions were vacuum dried and suspended in chloroform/methanol (9:1 [vol/vol]). The suspension was centrifuged, and the resulting white precipitate was solubilized in PBS for quantitative ELISA performed with the anti-GAG antibody. The purified polysaccharide (SGG; a kind gift of T. Fontaine, Institut Pasteur, Paris, France) was used for the preparation of standard curves and determination of polysaccharide concentrations in EV samples (48).
Protein composition of EVs. EV ultracentrifugation pellets were suspended in buffer containing 8 M urea-100 mM Tris (pH 7.5) for proteomic analysis. Protein samples were reduced with 5 mM dithiothreitol (DTT) for 30 min at 23°C and then alkylated with 20 mM iodoacetamide in the dark at room temperature, for 30 min. Subsequently, the endoproteinase LysC (Promega) was added for the first digestion step (protein-to-Lys-C ratio ϭ 80:1) for 3 h at 30°C. The sample was then diluted to reach a 1 M urea concentration with 100 mM Tris (pH 7.5), and trypsin (Promega) was added to the sample (protein-totrypsin ratio ϭ 50:1). The samples were digested for 16 h at 37°C. Proteolysis was stopped by the addition of 1% formic acid (FA). The resulting peptides were desalted using a Sep-Pak SPE cartridge (Waters) according to the manufacturer's instructions.
Analysis of digested peptides by the use of liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) was performed on an Orbitrap Q Exactive Plus mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) coupled to an EASY-nLC 1000 liquid chromatograph (Thermo Fisher Scientific). The peptides were loaded and separated at 250 nl · min Ϫ1 on a homemade C 18 50-cm capillary column with a picotip silica emitter (Dr. Maisch GmbH, Ammerbuch-Entringen, Germany) (75-m diameter filled with 1.9-m-pore-size Reprosil-Pur Basic C 18 -HD resin) equilibrated in solvent A (0.1% formic acid). The peptides were eluted using a gradient of solvent B (acetonitrile [ACN], 0.1% FA) of from 2% to 18% for 110 min, 18% to 30% for 35 min, and 30% to 45% for 15 min under conditions of a 250 nl/min flow rate. The total duration of the chromatographic run was 185 min, including high-ACN-level steps and column regeneration. Mass spectra were obtained in data-dependent acquisition mode with Xcalibur 2.2 software (Thermo Fisher Scientific, Bremen) with automatic switching between MS and MS/MS scans using a top-10 method. Spectral resolution corresponded to 70,000 (at m/z 400) with a target value of 3 ϫ 10 6 ions. The scan range was limited from 300 to 1,700 m/z. Peptide fragmentation was performed via higher-energy collision dissociation (HCD), with the energy set at a normalized collision energy (NCE) value of 28. The intensity threshold for the ion selection was set at 1 ϫ 10 6 ions, with charge exclusion settings of z ϭ 1 and Ͼ7. The MS/MS spectra were acquired at a resolution of 17,500 (at m/z 400). The isolation window was set at 1.6 Th. Dynamic exclusion was employed within 45 s.
A data search was performed with MaxQuant tool (76) (version 1.5.3.8) with the Andromeda search engine against the A. fumigatus A1163 database (9,942 entries, downloaded from https://www.uniprot .org [accessed 18 September 2019]). As search parameters, carbamidomethylation of cysteines was set as a fixed modification and oxidation of methionine and protein N-terminal acetylation were set as variable modifications. The mass tolerances in MS and MS/MS were set to 5 ppm and 20 ppm, respectively. The maximum peptide charge value was set to 7, and 7 amino acids were required as the minimum peptide length. A false-discovery rate of 1% was set for both protein and peptide levels. Four independent EV ultracentrifugation pellets were analyzed for protoplasts submitted either to cell wall-regenerating conditions or repressing conditions. Two EV samples of freshly purified protoplasts were analyzed. Proteins identified in at least two independent experiments were assigned for Gene Ontology (GO). GO term enrichment was performed using the GOtermfinder feature of the AspGD database (http://www.aspergillusgenome.org/cgi-bin/GO/goTermFinder). Enriched GO terms were summarized by removing redundancy using the REVIGO Web server, available at http://revigo.irb.hr/ (77). Protein identification was based on the detection of one representative peptide by mass spectrometry after protein digestion. For sample preparation and qualitative comparison, EV samples were obtained from cultures with similar inocula and volumes. Revigo outputs were viewed in R (v4.0.0) using the treemap package. Treemap construction was based exclusively on proteins shared by all replicates.
Statistical analysis. All statistical analyses were performed using GraphPad Prism 6 software (GraphPad Software Inc.). Data sets were tested for normal distribution using Shapiro-Wilk or Kolmogorov-Smirnov normality tests. In the cases in which the data passed the normality test (alpha ϭ 0.05), they were further analyzed using the unpaired Student's t test. Multiple data sets were further analyzed using ordinary one-way analysis of variance (ANOVA), followed by the Tukey's multiplecomparison test. When at least one data set was nonnormally distributed, multiple data sets were analyzed by the nonparametric Kruskal-Wallis test.
Data availability. Accession numbers for proteins described in this work are available in Table 1. We confirm that the data supporting the findings of this study are available within the article and its supplemental material. Additional data supporting our findings are available from J.R. or M.L.R. upon reasonable request.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.