PLB-985 Neutrophil-Like Cells as a Model To Study Aspergillus fumigatus Pathogenesis

ABSTRACT Fungal infections remain a major global concern. Emerging fungal pathogens and increasing rates of resistance mean that additional research efforts and resources must be allocated to advancing our understanding of fungal pathogenesis and developing new therapeutic interventions. Neutrophilic granulocytes are a major cell type involved in protection against the important fungal pathogen Aspergillus fumigatus, where they employ numerous defense mechanisms, including production of antimicrobial extracellular vesicles. A major drawback to work with neutrophils is the lack of a suitable cell line system for the study of fungal pathogenesis. To address this problem, we assessed the feasibility of using differentiated PLB-985 neutrophil-like cells as an in vitro model to study A. fumigatus infection. We find that dimethylformamide-differentiated PLB-985 cells provide a useful recapitulation of many aspects of A. fumigatus interactions with primary human polymorphonuclear leukocytes. We show that differentiated PLB-985 cells phagocytose fungal conidia and acidify conidia-containing phagolysosomes similar to primary neutrophils, release neutrophil extracellular traps, and also produce antifungal extracellular vesicles in response to infection. In addition, we provide an improved method for the isolation of extracellular vesicles produced during infection by employing a size exclusion chromatography-based approach. Advanced liquid chromatography-tandem mass spectrometry (LC-MS/MS) proteomics revealed an enrichment of extracellular vesicle marker proteins and a decrease of cytoplasmic proteins in extracellular vesicles isolated using this improved method. Ultimately, we find that differentiated PLB-985 cells can serve as a genetically tractable model to study many aspects of A. fumigatus pathogenesis. IMPORTANCE Polymorphonuclear leukocytes are an important defense against human fungal pathogens, yet our model systems to study this group of cells remain very limited in scope. In this study, we established that differentiated PLB-985 cells can serve as a model to recapitulate several important aspects of human polymorphonuclear leukocyte interactions with the important human fungal pathogen Aspergillus fumigatus. The proposed addition of a cultured neutrophil-like cell line to the experimental toolbox to study fungal pathogenesis will allow for a more mechanistic description of neutrophil antifungal biology. In addition, the easier handling of the cell line compared to primary human neutrophils allowed us to use PLB-985 cells to provide an improved method for isolation of neutrophil-derived extracellular vesicles using size exclusion chromatography. Together, these results provide significant tools and a baseline knowledge for the future study of neutrophil-derived extracellular vesicles in the laboratory.

ABSTRACT Fungal infections remain a major global concern. Emerging fungal pathogens and increasing rates of resistance mean that additional research efforts and resources must be allocated to advancing our understanding of fungal pathogenesis and developing new therapeutic interventions. Neutrophilic granulocytes are a major cell type involved in protection against the important fungal pathogen Aspergillus fumigatus, where they employ numerous defense mechanisms, including production of antimicrobial extracellular vesicles. A major drawback to work with neutrophils is the lack of a suitable cell line system for the study of fungal pathogenesis. To address this problem, we assessed the feasibility of using differentiated PLB-985 neutrophil-like cells as an in vitro model to study A. fumigatus infection. We find that dimethylformamide-differentiated PLB-985 cells provide a useful recapitulation of many aspects of A. fumigatus interactions with primary human polymorphonuclear leukocytes. We show that differentiated PLB-985 cells phagocytose fungal conidia and acidify conidia-containing phagolysosomes similar to primary neutrophils, release neutrophil extracellular traps, and also produce antifungal extracellular vesicles in response to infection. In addition, we provide an improved method for the isolation of extracellular vesicles produced during infection by employing a size exclusion chromatography-based approach. Advanced liquid chromatography-tandem mass spectrometry (LC-MS/MS) proteomics revealed an enrichment of extracellular vesicle marker proteins and a decrease of cytoplasmic proteins in extracellular vesicles isolated using this improved method. Ultimately, we find that differentiated PLB-985 cells can serve as a genetically tractable model to study many aspects of A. fumigatus pathogenesis. IMPORTANCE Polymorphonuclear leukocytes are an important defense against human fungal pathogens, yet our model systems to study this group of cells remain very limited in scope. In this study, we established that differentiated PLB-985 cells can serve as a model to recapitulate several important aspects of human polymorphonuclear leukocyte interactions with the important human fungal pathogen Aspergillus fumigatus. The proposed addition of a cultured neutrophil-like cell line to the experimental toolbox to study fungal pathogenesis will allow for a more mechanistic description of neutrophil antifungal biology. In addition, the easier handling of the cell line compared to primary human neutrophils allowed us to use PLB-985 cells to provide an improved method for isolation of neutrophil-derived extracellular vesicles using size exclusion chromatography. Together, these results provide significant tools and a baseline knowledge for the future study of neutrophil-derived extracellular vesicles in the laboratory. KEYWORDS Aspergillus fumigatus, PLB-985, phagocytosis, fungal pathogens, extracellular vesicles, HL-60 F ungal infections remain a tremendous source of global morbidity and mortality.
More than 1 billion individuals are affected per year, with invasive infections killing numbers comparable to other leading bacterial pathogens (.1.5 million per year [gaffi.org] [1]). Deadly invasive infections are caused by a relatively small number of fungi, with most of these attributed to members of the genera Candida, Pneumocystis, Cryptococcus, and Aspergillus (2). Aspergillus fumigatus is the major cause of aspergillosis and is particularly dangerous to immunocompromised individuals suffering from neutropenia (3). There is also emerging evidence to suggest that invasive aspergillosis may contribute to coronavirus disease 2019 (COVID-19)-related deaths (4,5), but challenges in safely obtaining bronchoalveolar lavage samples from these patients have often made confirmatory diagnosis difficult. Despite the obvious importance of A. fumigatus in the clinic, our understanding of this important pathogen remains lacking in many aspects, in part due to a lack of tractable experimental systems in the laboratory.
Mammals are continuously challenged by fungal pathogens. In fact, asexual spores of A. fumigatus, termed conidia, are thought to be inhaled by humans on a scale of hundreds per day (6). For most fungi, the body temperature of mammals is too high to allow for growth, but for organisms like A. fumigatus that thrive in compost piles at high temperatures, the human host provides fertile ground in the absence of a functional immune system (7). However, humans do have multiple additional defenses, including a mucociliary escalator to remove particles from the lungs, a robust epithelium, resident alveolar macrophages that eliminate the majority of the remaining fungal conidia, and infiltrating polymorphonuclear leukocytes (PMNs) that aid in clearance of conidia and destruction of fungal hyphae, among others (3,8). Neutrophils play an essential role in antifungal defense, due to their importance in killing fungal hyphae. This is well illustrated by the high susceptibility of neutropenic patients to A. fumigatus infections in clinics (reviewed in references 3 and 9).
Studies of primary human neutrophils have revealed the capacity of these cells to phagocytose conidia and release granules, neutrophil extracellular traps (NETs), and extracellular vesicles in response to invading pathogens (10)(11)(12). Phagocytosis occurs in conjunction with recognition of pathogen-associated molecular patterns by host pathogen recognition receptors. Studies of zebrafish and mice have shown that these internalized conidia can even be passed from neutrophils to macrophages for destruction of the fungus (13), implying complex intracellular trafficking. In neutrophils and macrophages however, internalized wild-type A. fumigatus conidia are capable of stalling phagolysosomal acidification to facilitate outgrowth (14).
In addition to phagocytosis, NETs are an important mechanism of defense against A. fumigatus that are produced in response to fungal recognition in a CD11b-dependent manner (11,15). NET production is most abundant against hyphae, but NETs are also sometimes produced in response to resting and swollen conidia (16). In vivo, NETs were shown to be present in mouse lungs during infection but were generally dispensable for fungal clearance (16,17), suggesting that alternative measures are required to eliminate fungal hyphae that escaped phagocytosis as conidia.
Neutrophilic granules and extracellular vesicles are two heterogeneous subcellular populations that also contribute widely to the antimicrobial response. Proteins associated with granules like lactoferrin are known to inhibit A. fumigatus conidial germination due to iron sequestration (18), and the abundant granule protein myeloperoxidase inhibits fungal hyphal growth (19). Extracellular vesicles were also shown to provide an antifungal mechanism against A. fumigatus hyphae and conidia (10). Antifungal extracellular vesicles produced by neutrophils in response to infection are capable of associating with the fungal cell wall and are in some cases internalized to deliver an unknown antifungal cargo. Intriguingly, the antifungal effect of these extracellular vesicles appeared to be tailored to the infecting pathogen, as spontaneously released vesicles and vesicles released in response to infection with a knockout strain deficient in production of the conidial pigment dihydroxynaphthalene (DHN)-melanin were not antifungal against wild-type fungus (10).
The molecular mechanisms behind these differences in antifungal potential are difficult to discern in the absence of a tractable model system for the genetic manipulation of neutrophils. Despite the importance of neutrophils in antifungal defense, the majority of studies are still performed in primary human PMNs isolated from venous blood, a resource that can be used only for small-scale studies and is unfortunately highly variable between blood donors. A laborious isolation procedure and limited half-life of about 19 h ex vivo (20) also make more mechanistic studies difficult. For these reasons, establishment of a cell culture line to study A. fumigatus pathogenesis in neutrophils or neutrophil-like cells would be highly advantageous. In recent years, some options to investigate specific aspects of neutrophil biology in culture have emerged (21). In particular, the myeloid cell line PLB-985, a derivative of HL-60 granulocytic cells, seems a tractable model to study NET formation in response to bacterial infection after differentiation into neutrophil-like cells (22,23). Of note, the related HL-60 cell line has also been used to deliver azole drugs to A. fumigatus-infected mice (24), but not to our knowledge in studies of A. fumigatus pathogenesis.
Here, we hypothesized that differentiated human PLB-985 neutrophil-like cells could serve as a model system to study the host pathogenesis of A. fumigatus in a blood donor-independent manner. We tested this hypothesis by defining the phagocytosis and intracellular trafficking of A. fumigatus conidia by PLB-985 cells, the production of neutrophil extracellular traps, and the release of extracellular vesicles produced in response to fungal infection. We find numerous similarities between this cell line and primary neutrophils that suggest differentiated PLB-985 cells can serve as a tractable in vitro model system for further study of some aspects of A. fumigatus pathogenesis.

RESULTS
Differentiated PLB-985 neutrophil-like cells phagocytose opsonized A. fumigatus conidia. We set out to assess the feasibility of using N,N-dimethylformamide (DMF)-differentiated PLB-985 (dPLB) cells as a model for neutrophil phagocytosis, NET production, and extracellular vesicle release in response to A. fumigatus infection. Upon infection, neutrophils are known to rapidly phagocytose and process A. fumigatus conidia. We therefore challenged dPLB cells with opsonized A. fumigatus resting conidia and assessed various parameters of infection.
First, we determined the phagocytic capacity of PLB-985 cells using confocal laser scanning microscopy following incubation of cells with fluorescein isothiocyanate (FITC)-labeled, opsonized conidia (green). We observed that dPLBs were capable of phagocytosis, as evidenced by a phagosomal membrane surrounding the conidia (Fig. 1A). A similar result was confirmed for primary human PMNs with this experimental setup (see Fig. S1A in the supplemental material). The dPLB cells only infrequently phagocytosed unopsonized conidia, highlighting the importance of opsonization and consistent with previous findings in primary human neutrophils (25). To better quantify phagocytosis, we analyzed 5,000 host cells postinfection with opsonized conidia by single-cell analysis using imaging flow cytometry. To do this, we counterstained with calcofluor white (CFW; blue) to elucidate nonphagocytosed conidia ( Fig. 1B and Fig. S1B and S1C). Quantification revealed that approximately 22% of wild-type and 26% of nonpigmented DpksP conidia were phagocytosed by dPLBs after 2 h, while undifferentiated PLB-985 cells exhibited limited phagocytosis of both wildtype and nonpigmented conidia (Fig. 1C). As expected, these phagocytosis percentages were lower than those of primary human PMNs, which aggressively phagocytosed conidia (Fig. S1D). To determine whether infection also resulted in an activated immune response, we measured proinflammatory cytokine levels by an enzyme-linked immunosorbent assay (ELISA). Coincubation of dPLB cells with wild-type spores resulted in production of significantly increased levels of interleukin 8 (IL-8), but not IL-1b after 24 h of coincubation, comparable to time points with primary human PMNs ( Fig. 1D and Fig. S2). Together, these results suggested that DMF-differentiated PLB-985 cells could serve as a suitable model for A. fumigatus pathogenesis and warranted further investigation.
Internalized A. fumigatus conidia are processed inside dPLB phagolysosomes. Following internalization of conidia by professional phagocytes like macrophages and neutrophils, conidia-containing phagosomes fuse with lysosomes to acidify the compartment and aid in fungal killing. Numerous additional enzymes are activated upon phagolysosomal acidification to degrade and digest the internalized fungal conidia. The acidification and maturation of phagolysosomes are regulated by protein complexes assembled on the phagolysosomal membrane, including lysosome-associated membrane protein 1 (LAMP-1), 2, and 3; several vacuolar ATPases (V-ATPases); Ras-related protein 5 (RAB5) and 7; flotillin 1 and 2; and numerous others (26). DHN-melanin on the surface of conidia can interfere with these processes in alveolar macrophages, monocytes, and primary neutrophils to delay fungal processing and phagolysosomal acidification (14). To test whether dPLB cells process conidia similarly to primary human PMNs, we first stained the cells with LAMP-1, a general endocytic marker on the membranes of phagolysosomes. LAMP-1 showed a clear signal around phagolysosomes of infected cells ( Fig. 2A), and the intensity of the signal did not differ between wild-type and DpksP conidia (Fig. 2B). Furthermore, loading of dPLB cells with Lysotracker, a weak base that becomes fluorescent under acidic conditions, showed that the DpksP conidia-containing phagosomes were significantly more acidified than those containing wild-type conidia, revealing that DHN-melanin can also block the acidification process in the dPLB model as well (Fig. 2C to D).
The interference of the acidification pathway in alveolar macrophages by fungal DHN-melanin occurs through the inhibition of V-ATPase assembly (27,28). This multiprotein complex plays a major role in lowering the pH from 6 to ,4.5 by pumping H 1 ions across the phagolysosomal membrane. Staining of the V-ATPase V1 subunit in dPLB cells revealed that the percentage of recruitment to conidia-containing phagolysosomes was fairly similar for wild-type and DpksP conidia ( Fig. S3A and S3B), suggesting that other proton pumps may contribute to acidification of these phagolysosomes in neutrophil-like cells. An additional defense mechanism for intraphagosomal degradation of pathogens occurs via reactive oxygen species (ROS) production. As expected, we observed dPLB cells to be capable of producing ROS after staining with CellROX Orange (Fig. S3C), consistent with the literature (29,30).
A. fumigatus triggers formation of NETs from dPLB cells. One defense mechanism employed by neutrophilic granulocytes to fight against pathogens is the formation of NETs. These structures consist of condensed chromatin and various enzymes that are released into the extracellular space and typically correlate with death of the cell (31). NETs contain several antifungal proteins, which are responsible for the fungistatic effect against A. fumigatus hyphae (16). To test whether dPLBs are also able to form NETs in response to A. fumigatus and serve as a model for this pathway, we coincubated dPLB cells with hyphae and stained for nucleic acid with 49,6-diamidino-2phenylindole (DAPI) (Fig. 3A). For a positive control, we induced NET formation using phorbol myristate acetate, a known trigger of NET formation in PLB-985 cells (Fig. 3B) (22,32). We observed the presence of histone H3 embedded in the DNA fibers and an association of NETs with fungal hyphae (Fig. 3B). Taken together, these results showed that A. fumigatus is able to trigger NET formation in dPLB cells and that these NETs are specifically directed against the hyphae.
dPLB cells produce extracellular vesicles in response to infection. Recently, a new defense mechanism from neutrophilic granulocytes was discovered, the production of antimicrobial extracellular vesicles (33,34). These small lipid-enclosed nanoparticles are released from primary PMNs after contact with microorganisms and, in the case of A. fumigatus conidia, can inhibit fungal growth after coincubation (10). This effect is likely due in part to their protein cargo, which consists of antimicrobial peptides such as myeloperoxidase, azurocidin, and cathepsin G. To assess whether dPLB cells can produce extracellular vesicles spontaneously or in response to A. fumigatus, we incubated cells with or without conidia for 2 and 4 h. Next, we isolated extracellular vesicles using two different methods; first using a previously described differential centrifugation-based approach (DC) (10) that enriches for medium-sized extracellular vesicles and a second approach that relies on size exclusion chromatography (SEC) to purify a more selective population of smaller extracellular vesicles. Using these methods, dPLB cells were observed to actively secrete extracellular vesicles over time in a manner comparable to primary human PMNs ( Fig. 4A and B). Using nanoparticle tracking analysis, we observed the median size of particles to be around 200 nm, similar to extracellular vesicles derived from primary neutrophils, a feature that was comparable between infection-derived and spontaneously released extracellular vesicles (Fig. 4C to D).
We next compared the protein content of dPLB-derived extracellular vesicles isolated at the 2-h time point using each isolation method with and without fungal infection by LC-MS/MS-based proteomics analysis. We were able to identify 1,984 unique proteins across all four samples (see Data Set S1 in the supplemental material). The majority of identified proteins (737 proteins) were found in all four samples (Fig. 5A). We expected that isolation using size exclusion chromatography would improve the quality of the isolated particles as has been shown previously (35), and this was in fact the case. We observed an increase in extracellular vesicle markers like the CD63 and CD81 tetraspanins and tumor susceptibility gene 101 (TSG101), and a decrease in cytoplasmic proteins like calnexin (CANX) ( Table 1). Extracellular vesicles produced by primary PMNs in response to A. fumigatus infection are known to be enriched for antimicrobial cargo proteins like azurocidin, cathepsin G, and defensin (10). In line with the primary neutrophil data, we found that dPLBs also contained many of the same proteins in both infection-derived and spontaneously released extracellular vesicles (Table 1), as evidenced by an UpSetR plot showing overlapping protein cohorts (Fig. S4) (10), albeit with some exceptions. For example, dPLBs appeared to lack defensin, neutrophil elastase, and some histone proteins previously observed. Ultimately, improvements in LC-MS/MS technology and the advantage of higher input amounts of extracellular vesicle protein using large amounts of dPLB cells in culture resulted in significantly more proteins detected here compared to efforts using primary neutrophils (10), proving that dPLBs offer a scalable system for the elucidation of novel mechanisms of neutrophil extracellular vesicle biology.
dPLB extracellular vesicles produced against A. fumigatus limit fungal growth. The most compelling feature of infection-derived extracellular vesicles of human neutrophils is likely their antifungal capacity, as we previously reported (10). We set out to determine whether the extracellular vesicles produced by dPLB cells in response to opsonized A. fumigatus conidia are antifungal to a mitochondrial-GFP (green fluorescent protein) reporter strain used previously as a marker of fungal viability (36). First, we incubated the reporter strain with dPLB cells and primary human PMNs and assessed the capacity of the cells to control infection. PMNs were capable of minimizing fungal outgrowth after 22 h, whereas dPLBs did not completely contain fungal growth (Fig. 6A). We then assessed the antifungal capacity of the dPLB extracellular vesicles. For this experiment, extracellular vesicles from equal numbers of dPLBs and primary human PMNs were isolated using differential centrifugation and found to be antifungal against the mitochondrial-GFP reporter strain. This was evidenced by fragmentation of fungal mitochondria after administration of extracellular vesicles to germinating conidia, which were allowed to germinate for 6 h prior to overnight incubation with extracellular vesicles, 3 mM H 2 O 2 as a positive control, or left untreated as a negative control (Fig. 6B). Although the mitochondrial reporter clearly indicated fragmentation, the extracellular vesicles from dPLBs did appear to be in some cases less effective in limiting the length of hyphae than those from primary PMNs. It is important to note that dPLBs produced slightly fewer extracellular vesicles, which could partially explain the decreased antifungal activity in these assays (Fig. 6B). Additional representative images of dPLB-derived extracellular vesicles are shown to indicate the spectrum of phenotypes observed for each experimental condition (Fig. S5). Collectively, these results suggest that dPLB infectionderived extracellular vesicles can limit A. fumigatus hyphal growth, similar to primary human PMNs (10). and SEC-based isolation (C). Input data included values from all replicates using the RMN data included in Data Set S1 in the supplemental material. Plots were created using ggplot2 in R. Proteomics data are from three analytical replicates of three independent biological replicates. Orange circles represent proteins with greater than twofold change and P value of #0.05. Selected proteins are named for clarity.

DISCUSSION
Neutrophils are critical players in the immune response to fungal pathogens stemming from their numerous antimicrobial capacities (10,37). To study neutrophil functions, numerous cell lines that model these abilities in vitro have been considered with limited success. The most promising are the HL-60 leukemia cell line and PLB-985 sublineage, which have been used as models for aspergillosis and bacterial killing (22)(23)(24)29). In this study, we analyzed different aspects of the interaction between DMF-differentiated PLB-985 myeloid cells and the fungus A. fumigatus for the first time, making frequent comparisons with data from primary neutrophils.
Primary neutrophils are capable of phagocytosing and processing A. fumigatus conidia, features that were maintained in dPLB cells. As expected for cultured cells, the observed phagocytosis rate of A. fumigatus conidia by dPLB cells was lower than that of primary human PMNs. It is possible that additional differentiation methods will ultimately reveal higher rates of phagocytosis, as PLB-985 cells differentiated with DMF are known to exhibit lower phagocytic activity than cells differentiated with dimethyl sulfoxide (38). In terms of cytokine release, we observed increased release of IL-8 from dPLBs after infection with opsonized A. fumigatus conidia, but not IL-1b, consistent with our observations of human PMNs. These results are also in agreement with those collected from primary PMNs coincubated with L-ficolin-opsonized conidia for 8 h, where secretion of IL-8 was also much higher than secretion of IL-1b (39).
We next stained dPLB cells for the endosomal marker LAMP-1 and measured the acidification of the phagolysosomes via Lysotracker to assess intracellular processing of fungal conidia. These experiments confirmed that around 20% of wild-type conidia are completely internalized inside the phagosomes and that lysosomal fusion occurs as a next step of conidial processing. Interestingly, and as already demonstrated for primary monocytes and neutrophils (14), melanized conidia showed less acidification around the phagolysosomal membrane than the mutant lacking DHN-melanin. This effect in monocytes is caused by inhibition of V-ATPase assembly. Surprisingly, staining for the V1 subunit of the V-ATPase did not reveal a significant difference between the wild-type and DpksP strains in dPLB cells as has been observed in other cell systems. These results suggest that neutrophils potentially employ additional mechanisms to acidify phagolysosomes in neutrophilic granulocytes after A. fumigatus infection or that the methods used were not sufficiently sensitive to detect slight differences in acidification. Together, these results suggest that dPLBs can serve as an intriguing model to study various aspects of neutrophil phagocytosis and intracellular processing in a more tractable in vitro model system.
A conidium that escapes phagocytosis can sometimes germinate into a hypha, a morphotype that is much more difficult to eliminate and requires the antifungal activity of neutrophils. One mechanism that aids in control of A. fumigatus hyphae is the release of fungistatic NETs composed of DNA and antimicrobial proteins. By staining for extracellular DNA and NET constituent histone H3, we could clearly show that dPLB cells also produce NETs in response to A. fumigatus hyphae, comparable to primary human PMNs and in agreement with other studies challenging dPLBs with bacterial pathogens (22,23). Production of extracellular vesicles from primary neutrophils coincubated with Aspergillus conidia or bacteria like Staphylococcus aureus is an important defense

(A) Conidia from
A. fumigatus strain AfS35 containing plasmid pJW103 expressing a mitochondrial GFP reporter (green) were opsonized and coincubated with freshly harvested human PMNs or dPLBs. After overnight incubation (22 h), samples were stained with CFW for 10 min, and images were taken using a Zeiss LSM 780 confocal microscope. Bars, 20 mm. (B) The A. fumigatus strain AfS35 expressing a mitochondrial GFP reporter (green) was first grown for 6 h and then incubated overnight with spontaneously released extracellular vesicles or infection-derived extracellular vesicles isolated from primary human PMNs or dPLBs. Fungal hyphae were then stained with CFW. For a control, untreated hyphae and hyphae treated with 3 mM H 2 O 2 to induce cell death are included. An intact mitochondrial network is shown by a filamentous network, whereas a disrupted network is shown by fragmentation or the lack of green signal. Bars, 5 mm. strategy against microorganisms (10,33). We found that dPLB cells were also able to generate comparable populations of extracellular vesicles to PMNs upon contact with opsonized conidia. Accumulation of extracellular vesicles of approximately 200 nm in diameter was observed over time, independently of the isolation method we applied. Proteomics analysis revealed that by using a size exclusion chromatography-based isolation approach, the extracellular vesicle population could be enriched for extracellular vesicle marker proteins like CD63, CD81, and TSG101. We observed that the samples obtained by this method had a higher abundance of proteins and showed more differences between the spontaneously released and infection-derived extracellular vesicles. In particular, infection-derived extracellular vesicles were enriched in immunoglobulins and complement factors compared to spontaneously released vesicles. These serum proteins may be introduced to the infection system during opsonization of the fungal conidia despite thorough washing. It remains unclear whether these immunoglobulins and complement factors play an important role in extracellular vesicle biology during A. fumigatus infection. As a hint for such a role, the association of immunoglobulins and complement proteins was shown previously to be important in systemic lupus erythematosus (SLE) disease, where extracellular vesicles with immunoglobulin cargo acted as immune complexes to mediate inflammation (40,41).
In the proteomics analysis of extracellular vesicles derived from primary neutrophils infected with A. fumigatus, several antimicrobial peptides, such as cathepsin G and azurocidin, were detected (10). Coincubation of these extracellular vesicles from primary neutrophils with fungal hyphae arrested growth, which we confirmed here. Using dPLB cells, we identified extracellular vesicles containing cathepsin G and azurocidin; however, the distribution remained relatively equal between spontaneously released and infectionderived extracellular vesicles. We suspect that this is due to the differentiation procedure of dPLB cells with DMF, which provides a mild inflammatory stimulus. Nevertheless, we observed mitochondrial damage in hyphae treated with infection-derived but not spontaneously released extracellular vesicles. We can postulate several reasons for this result. First, it is possible that immunoglobulins are playing a role in targeting extracellular vesicles to fungal hyphae. In this case, the protein cargo may be the same, but targeting would be inefficient in spontaneously released extracellular vesicles. Of note, opsonization of the bacterial pathogen Staphylococcus aureus was not found to influence the antimicrobial capacity of neutrophil-derived extracellular vesicles (33). A second option is that additional extracellular vesicle cargo molecules such as RNA or lipids may play a role in the antifungal activity. It has been shown in many instances that extracellular vesicles can contain small RNAs that exert distinct functions during infection, especially in regard to plant fungal pathogens (42,43). We believe that a combination of factors is likely involved in the antifungal effect of extracellular vesicles.
Part of the explanation may also come from similarities between neutrophilic granules and extracellular vesicles. Granules and extracellular vesicles are distinct cellular features, but they do share some biogenesis pathways and cargo molecules. For example, the azurophilic granules contain myeloperoxidase to aid in pathogen killing in the endocytic pathway via fusion of the granules to pathogen-containing phagosomes (44), but myeloperoxidase is also released in a CD63-negative population of extracellular vesicles known as microvesicles (45). Based on the proteomics analysis of this study, as well as our previous work (10), it seems possible that microvesicle cargo proteins like myeloperoxidase contribute at least in part to the antifungal activity. However, the presence of myeloperoxidase in spontaneously released extracellular vesicles and the lack of susceptibility of patients lacking myeloperoxidase to fungal infections (46) hint that other factors are also involved. Besides, the shared marker protein CD63 nicely highlights another link between granules and extracellular vesicles, the shared biogenesis from multivesicular bodies of azurophilic granules and the extracellular vesicle subset called exosomes (47).
In conclusion, our results suggest that DMF-differentiated PLB-985 cells can be used as a model to study aspects of the interaction of the human-pathogenic fungus A. fumigatus with neutrophilic granulocytes. Although they will never substitute for all experiments with neutrophils, we do believe that this system will serve as a useful tool for the genetic dissection of A. fumigatus pathogenesis in the future.

MATERIALS AND METHODS
Fungal strain cultivation and opsonization. Cultivation of A. fumigatus strains CEA10 (Fungal Genetics Stock Center; A1163), CEA17 DakuB ku80 DpksP (48), and AfS35/pJW103 (36) was performed on malt agar plates (Sigma-Aldrich) supplemented to a final concentration of 2% (wt/vol) agar for 5 days at 37°C. Conidia were harvested in sterile, ultrafiltrated water filtered through a 30-mm pore filter (MACS, Miltenyi Biotec). Prior to confrontation with PLB-985 cells, conidia were opsonized with normal human serum (Merck Millipore). Briefly, 900 ml of spore suspension was mixed with 100 ml of normal human serum and incubated in a thermomixer at 37°C for 30 min shaking at 500 rpm. The spore suspension was washed three times by collecting spores via centrifugation at 1,800 Â g for 1 min at 4°C and resuspending in fresh phosphate-buffered saline (PBS). Following washing, spores were enumerated in a Thoma chamber in preparation for infection assays.
Isolation of human PMNs. Fresh venous blood was drawn from adult healthy volunteers, aged 20 to 35 years, after obtaining written consent. This study was approved by the Jena University Hospital Institutional Review Board (approval number 5074-02/17) in agreement with the Declaration of Helsinki. Blood was collected in EDTA BD Vacutainer tubes (BD Biosciences) and mixed with PBS. Six milliliters of this solution was carefully layered onto (avoiding any mixing) 5 ml (1:1 [vol/vol]) of PolymorphPrep solution (Progen) and centrifuged 50 min at 500 Â g for gradient separation. After centrifugation, the neutrophil layer was collected, mixed with an equal volume of 0.5% PBS, and recentrifuged for 15 min at 400 Â g. To limit contamination of red blood cells, the pellet was lysed with the ACK lysis buffer (Gibco) and centrifuged as before. The PMNs were finally resuspended in RPMI medium without phenol red (Thermo Fisher Scientific) and after addition of trypan blue counted with a Luna automated cell counter (Logos Biosystem). For each experiment, the viability of neutrophils was $95%.
Phagocytosis assays. To assess phagocytic ability, we used a combination of imaging flow cytometry and confocal fluorescence microscopy. For both methods, the conidia were first stained with fluorescein isothiocyanate (FITC) and then after confrontation of dPLBs or primary neutrophils, counterstained with calcofluor white (CFW; Sigma-Aldrich). The FITC solution was obtained by dissolving FITC powder (Sigma-Aldrich) in 5 ml of 0.1 M sodium carbonate (Na 2 CO 3 ), followed by filtration through a 0.22-mmpore-size filter (Carl Roth). Afterwards 1 ml of this solution was mixed with 500 ml of spore suspension and incubated for 30 min at 37°C while shaking at 1,000 rpm in the dark. The spores were then pelleted and washed three times with PBS with 0.001% (vol/vol) Tween 20 (Carl Roth). During the last washing step, Tween was removed to avoid residual detergent in the samples (14,49). Before counting, the conidia were opsonized following the protocol described above. Spores were then added to cells and coincubated for 2 and 4 h at 37°C with 5% (vol/vol) CO 2. At each time point, CFW was added to a final concentration of 1 mg/ml and incubated for 1 min. Afterwards the cell suspension was transferred to microcentrifuge tubes, centrifuged at 600 Â g for 2 min at 4°C, and washed with PBS twice. After the supernatant was discarded, the pellet was fixed with 150 ml of 3.7% (vol/vol) formaldehyde in PBS at room temperature for 15 min and subsequently washed as described above. For imaging flow cytometry measurements, the cells were resuspended in 150 ml of PBS and analyzed immediately or stored at 4°C for no more than 24 h. Four independent experiments were performed, and for each replicate, 5,000 cells were analyzed using the ImageStream X Mark II (Luminex). A 488-nm laser was used to detect FITC staining, and a 405-nm laser was used to detect CFW staining. The laser voltage was adjusted depending on the sample. For wild-type conidia, we used 10 mW of 488-nm laser and 2 mW for the 405-nm laser, and for DpksP conidia, we used 1 and 2 mW of the respective lasers. The fluorescence intensities of the samples were compensated and analyzed with the IDEAS software (Luminex). An example of the gating strategy used for analysis is provided (see Fig. S1B and S1C in the supplemental material). For the fluorescence microscopy analysis, cells were seeded in 8-well m-slides (Ibidi) and visualized using a Zeiss LSM 780 confocal microscope (Carl Zeiss) from at least three biological replicates.
Immunofluorescence assays. For immunofluorescence assays, cells were allowed to adhere in a 24well plate with poly-L-lysine-coated glass coverslips for 1 h (Merck). After infection with labeled A. fumigatus spores and incubation for the noted times, cells were fixed using 3.7% (vol/vol) formaldehyde in PBS for 10 min, rinsed three times with PBS, permeabilized for 15 min using 0.1% (vol/vol) Triton X-100 in PBS or 0.1% (wt/vol) saponin in PBS, and then blocked for 30 min with 2% (wt/vol) bovine serum albumin (BSA). After permeabilization, washed cells were incubated with primary rabbit anti-Lamp-1 antibody (Abcam ab24170; 1:100 dilution), anti-V-ATPase V1 subunit antibody (Abcam ab73404; 1:100 dilution), or anti-histone H3 (Cell Signaling DIH2; 1:200 dilution) antibody in 1% (wt/vol) BSA in PBS, Database search and data analysis. Tandem mass spectra were searched against the UniProt database of Homo sapiens (https://www.uniprot.org/proteomes/UP000005640; 17 May 2021) and Neosartorya fumigata (https://www.uniprot.org/proteomes/UP000002530; 17 May 2021) using Proteome Discoverer (PD) 2.4 (Thermo) and the algorithms of Mascot 2.4.1 (Matrix Science), Sequest HT (version of PD2.2), MS Amanda 2.0, and MS Fragger 3.2. Two missed cleavages were allowed for tryptic digestion. The precursor mass tolerance was set at 10 ppm, and the fragment mass tolerance was set at 0.02 Da. Modifications were defined as dynamic Met oxidation and protein N-terminal acetylation with and without methionine loss as well as static Cys carbamidomethylation. A strict false discovery rate (FDR) of ,1% (peptide and protein level) were required for positive protein hits. If only one peptide per protein has been identified, the hit was accepted if the Mascot score was .30 or the MS Amanda score was .300 or the Sequest score was .4 or the MS Fragger score was .8. The Percolator node of PD2.4 and a reverse decoy database was used for q value validation of spectral matches. Only rank 1 proteins and peptides of the top scored proteins were counted. Label-free protein quantification was based on the Minora algorithm of PD2.2 using a signal-to-noise ratio of .5. Imputation of missing quantification values was applied by setting the abundance to 75% of the lowest abundance identified for each sample. Normalization was based on a replicate median total peptide sum approach, which was calculated based on the sum of all identified peptide abundance values per replicate sample. The sums of each of the three replicates from the four sample groups were used to calculate median values. Normalization factors were calculated by dividing median values of the respective sample group by the abundance sum of each sample. Normalization factors were multiplied with single protein abundance values of each replicate/sample. The P values are based on a Student`s t test. Ratio-adjusted P values were calculated by dividing P values with the log 4 ratio of the protein abundance levels. Significant differences in protein abundance were defined when the following three requirements were reached: at least a fourfold change in abundance (up and down), a ratio-adjusted P value of ,0.05, and identified in at least two of three replicates of the sample group with the highest abundance. Intersection plots were created using the UpSetR package (51) and include only proteins that were detected in at least two replicates of a given sample. Volcano plots were created using ggplot2 in R using the replicate median total peptide sum normalized (RMN) data for all proteins detected in Data Set S1 in the supplemental material.
Fungal mitochondrial reporter inhibition assay. To characterize growth inhibition and killing of hyphae by fungus-induced extracellular vesicles, an A. fumigatus strain expressing a mitochondrial GFP reporter (AfS35/pJW103 [36]) has been used as described previously (10). Extracellular vesicles produced by 1 Â 10 7 dPLB cells or PMNs were isolated and coincubated with the reporter strain. After 16 h of incubation, samples were stained with CFW, and images were taken using Zeiss LSM 780 confocal microscope (Carl Zeiss).
Statistical analysis. Data were plotted and statistically analyzed using GraphPad Prism software 5.0 (GraphPad Software) unless otherwise noted. The Student's t test was used for significance testing when comparing two groups. Differences between the groups were considered significant at a P value of ,0.05. Throughout the article, significance is denoted as follows: *, P # 0.05; **, P # 0.01, ***, P # 0.001; ns, not significant.
Ethics statement. Peripheral human blood was collected from healthy volunteers only after written informed consent was provided. The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the University Hospital Jena (approval number 5074-02/17).
Data availability. The mass spectrometry proteomics data have been deposited to the Proteome-Xchange Consortium via the PRIDE partner repository (52) with the data set identifier PXD027032.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. DATA SET S1, XLSX file, 18.9 MB.