A phylogenetic approach to explore the Aspergillus fumigatus conidial surface-associated proteome and its role in pathogenesis

Abstract Aspergillus fumigatus , an important pulmonary fungal pathogen causing several diseases collectively called aspergillosis, relies on asexual spores (conidia) for initiating host infection. Here, we used a phylogenomic approach to compare proteins in the conidial surface of A. fumigatus , two closely related non-pathogenic species, Aspergillus fischeri and Aspergillus oerlinghausenensis , and the cryptic pathogen Aspergillus lentulus . After identifying 62 proteins uniquely expressed on the A. fumigatus conidial surface, we assessed null mutants for 42 genes encoding conidial proteins. Deletion of 33 of these genes altered susceptibility to macrophage killing, penetration and damage to epithelial cells, and cytokine production. Notably, a gene that encodes glycosylasparaginase, which modulates levels of the host pro-inflammatory cytokine IL-1β, is important for infection in an immunocompetent murine model of fungal disease. These results suggest that A. fumigatus conidial surface proteins and effectors are important for evasion and modulation of the immune response at the onset of fungal infection.


Introduction
Pulmonary fungal diseases caused by the environmental mold Aspergillus fumigatus lead to signi cant morbidity and mortality.In patients with weakened lung defenses arising from immunosuppression, a chronic respiratory condition, or a prior respiratory infection, asexual spores or conidia of A. fumigatus can evade the lung defenses, germinate, and cause disease 1 .However, our understanding of what makes A. fumigatus a successful pathogen compared to other species remains incomplete.Current evidence suggests that several different Aspergillus species, including A. fumigatus, likely independently evolved the ability to cause human disease 2 (Fig. 1A), raising the hypothesis that species-speci c genes contribute to disease.
The early stages of disease-marked by the interaction between the inhaled conidia and the host-may prove insightful for unraveling A. fumigatus pathogenicity.Among studies of other microbial pathogens, there has been an increasing focus on conidial surface protein characterization since they mediate the rst encounter with the host immune system.The conidial cell wall comprises a β-1,3-glucan and chitin body, which is covered by rodlet and melanin layers, where proteins are anchored 3,4 .These and other surface proteins play pivotal roles in morphogenesis, resistance to environmental stressors, substrate adherence, and virulence 5 .The hydrophobin RodA is a signi cant component of the rodlet layer and is essential for cell wall physical resistance and permeability, preventing immune recognition 6,7 .Another conidial surface protein, CcpA, is essential to maintain the correct surface structure and prevent immune recognition 8 .Recently, HscA has been demonstrated to anchor human p11 on phagosomal membranes, rewiring the vesicular tra cking to the non-degradative pathway, allowing the escape of conidia 9 .Other studies aimed to identify potential vaccine candidates, allergens, and biomarkers for diagnosis or characterization of A. fumigatus cell wall dynamics under different stresses or biological processes 10,11,12 .
Despite the importance of the conidia for disease, a comprehensive examination of conidial surface proteins and their contribution to pathogenicity in A. fumigatus and closely related species is lacking.Here, we analyzed the conidial surface proteome of A. fumigatus and three closely related species: (i) Aspergillus lentulus, which is a cryptic pathogen and close relative of A. fumigatus 13 ; (ii) Aspergillus oerlinghausenensis, the closest known relative of A. fumigatus, which is azole-resistant but nonpathogenic 14,15 ; and (iii) Aspergillus scheri, a close relative of A. fumigatus that is less virulent in several animal models and rarely causes disease in humans 16 (Fig. 1a).

Results
A. fumigatus has increased virulence in a chemotherapeutic murine model of Invasive Pulmonary Aspergillosis (IPA) and elicits higher cytokine production by bone marrow-derived macrophages (BMDMs).To understand whether the four species have in vivo and in vitro differences in pathogenicity, we comparatively analyzed the capacity of A. fumigatus, A. scheri (A ), A. oerlinghausensis (Aoe), and A. lentulus (Ale) to infect macrophages and susceptible hosts.A. fumigatus conidia are less engulfed and killed by Bone Marrow derived macrophages (BMDMs) than A , Aoe, and Ale (Figs. 1b and 1c). A. fumigatus conidia are recognized by macrophages and are intracellularly degraded by the endocytic pathway through the fusion of conidia-containing phagosomes and lysosomes, forming an acidic phagolysosome (PL) 17 .Melanin can facilitate evading this mechanism, as the A. fumigatus ΔpksP mutant (which lacks a key enzyme involved in melanin biosynthesis) induces higher levels of PL acidi cation than the corresponding wild-type strain ATCC46645 (Fig. 1d).However, there are no differences in PL acidi cation in A. fumigatus, A , Aoer, and Ale, indicating that these four species' mechanisms of inhibiting PL acidi cation are conserved (Fig. 1d).In contrast, A. fumigatus conidia induced higher levels of four cytokines-TNF-α, IL-6, IL-1β, and 1L-18-in BMDMs than the other three species (Figs.1e to 1h).Interestingly, A. fumigatus and Ale uniquely induced production of in ammasome cytokines IL-1β and IL-18, while A and Aoer conidia led to the production of lower levels of TNF-α than A. fumigatus and A conidia (Figs.1e to 1h).
Virulence of each species was assessed in a clinically relevant chemotherapeutic BALB/c murine model of IPA.The A. fumigatus strain was signi cantly more virulent than A , Aoe, and Ale strains (p-value > 0.005, Fig. 1i) and killed all mice after 5 days post-infection (d.p.i.).However, while there was no statistical difference among the other species, an increased virulence caused by Ale strain was observed when compared to the Aoe and A strains (40% vs. 50-60% of survival at 15 d.p.i., Fig. 1i).
Taken together, these results strongly indicate that A. fumigatus is more virulent in a chemotherapeutic murine model of IPA, is more e ciently recognized by BMDMs, and can elicit higher levels of cytokine production than the other three species.
The conidial surfome of A. fumigatus contains 62 unique proteins.Variation in host recognition observed between A. fumigatus and the other three species raises the hypothesis that conidial surface proteins may underlie differences in host recognition variation.To test this hypothesis, we conducted a trypsin-shaving proteomic analysis and identi ed A. fumigatus unique proteins on the conidial surface.To do so, samples of resting (0h) and swollen (4h, 37°C) conidia from the four Aspergillus species included in the study (A.fumigatus, A , Aoe, and Ale; Fig. 2a) were collected and processed for LC-MS/MS analysis as described in 5 (Fig. 2b).A total of 354 proteins were identi ed in A. fumigatus conidial surface [193 in resting and 161 in swollen conidia, and 122 (52.6%) shared between the two conditions].Similarly, in A , 594 proteins were identi ed as part of the conidial surfome [309 in resting and 285 in swollen conidia, and 204 (52.3%) shared between the two conditions].For Ale, we identi ed 1,027 conidial surface proteins [285 in resting and 742 in swollen conidia, but only 240 (30.1%) shared between the two conditions].Finally, 763 proteins were identi ed in Aoe conidial surface [393 in resting and 370 in swollen conidia, and 269 (54.5%) shared between the two conditions] (Supplementary Figures S1; Supplementary Table S1).
Comparative analyses of the surface-associated proteome or "surfome" revealed 62 conidial surface proteins were A. fumigatus-speci c, i.e., only detected in A. fumigatus conidia.Fifty-six of the genes encoding these proteins are shared by the majority of 206 A. fumigatus isolates, whereas six (AFUB_028320, AFUB_039180, AFUB_045200, AFUB_050510, AFUB_069520, and AFUB_094680) were more variable, belonging to the accessory pangenome (pangenome information from 18 ).In the other three Aspergillus species, 53, 118, and 385 conidial proteins were unique to A , Aoer, and Ale, respectively.Numerous conidial proteins were also broadly shared; 188 proteins were shared by at least two species, 146 were shared by at least three species, and 75 were shared by all four species (Fig. 2c).
Functional categorization of the A. fumigatus-speci c conidial surfome was carried out manually according to the information available in FungiDB database (https://fungidb.org/fungidb/app/).Most genes encode enzymes involved in many biological processes such as cell wall modi cation, metabolism, cell signaling, and secondary metabolite biosynthesis.Some of these enzymes have previously been identi ed as allergens and proteases (Fig. 2c; Table 1).Other surface proteins belonged to less enriched functional categories, such as structural proteins and transporters.However, more than a third of the proteins identi ed as part of the A. fumigatus-speci c conidial surfome had unknown functions (Fig. 2c).To further explore the putative function of identi ed proteins, we predicted the presence of secretion and GPI-anchor peptides using SignalP − 5.0 (https://services.healthtech.dtu.dk/service.php?SignalP-5.0) and PredGPI (http://gpcr2.biocomp.unibo.it/gpipe/index.htm).Thirty-one out of the 62 proteins are predicted to have a signal peptide cleavage site, while only two have high probability (> 99%) of harboring a GPI-anchoring signal (Table 1).One protein (encoded by Afu3g13755) did not have a homologue in the A. fumigatus A1163 background strain and was not examined further.We examined the expression of these 62 genes in a recently published RNAseq dataset that describes gene modulation during conidial germination 19 (Fig. 2d).About 27% of the surfome genes (16 genes) do not change their gene expression during the rst 16 h of germination while 33% (21 genes) have increased expression during the 2 to 16 h germination window (Fig. 2d).About 24% of the surfome genes (15 genes) are expressed late, while 16% (10 genes) are expressed early (Fig. 2d).
Further characterization of A. fumigatus-speci c conidial surface proteins was carried out using genedeletion strains and phenotypic assays.To do so, we accessed strains from the COFUN homozygous deletion library for 42 genes identi ed in our previous proteomic analysis following the high-throughput methodology described in 20 .After three attempts, we were unable to construct deletion mutants for the reminaing 20 of the 62 A. fumigatus-speci c genes that were not in the library.It remains to be determined if we were not able to delete the 20 remaining genes because they encode essential proteins.
Deletions were PCR validated (primers listed in Supplementary Table S2).An exhaustive phenotypic screen was carried out for each deletion mutant to identify the involvement of the A. fumigatus surfome in diverse bioprocesses related to virulence (e.g., growth at the temperature of the human body).Growth rates in solid MM at 37°C revealed a defect in growth for ΔAFUB_058080 mutant strain, while ΔAFUB_047280 displayed increased or faster growth compared to the parental A1160 strain.None of the mutants exhibited increased sensitivity to heat (44°C) (Fig. 2e and Table 1).However, ΔAFUB_058080 was found to be more sensitive than the A1160 wild-type strain to the cell wall stressor Congo Red (CR).Similarly, ΔAFUB_040910 and ΔAFUB_100920 showed decreased and increased susceptibility to oxidative stress, respectively (Fig. 2e and Table 1).Germination rates for all the mutants and wild-type strain were also determined; most null mutant strains (33/42) behaved similarly to A1160 reference strain.The remaining deletion strains could be differentiated into two groups: i) slow-germinating (below 30% of A1160 germination rate; ΔAFUB_035440, ΔAFUB_057080, ΔAFUB_058080, ΔAFUB_100460) and ii) fast-germinating (above 30% of A1160 germination rate; ΔAFUB_010730, ΔAFUB_039800, ΔAFUB_040910; ΔAFUB_043810, ΔAFUB_095010) (Fig. 2f and Table 1).
The crystal violet (CV) assay was used as an indirect measurement of the adhesion properties of conidia at both resting and swollen stages for all the strains.Interestingly, signi cant differences between the wild-type and deletion strains were only observed in two instances: ΔAFUB_000410 displayed reduced adhesion properties in comparison with A1160, while ΔAFUB_041890 showed a higher capacity to adhere to surfaces than the wild-type strain (Fig. 2g and Table 1).
Taken together, these results suggest that A. fumigatus conidial surface-associated proteins are also involved in the correct assembly of the cell wall and have an important role in combatting stress and adhering to surfaces.
A subset of A. fumigatus surfome proteins is involved in host-pathogen interactions.To understand the role of the A. fumigatus surfome in the establishment of infection, we individually assessed the role of each of the 42 A. fumigatus-speci c surfome proteins in survival and tissue invasion using in vitro models of infection.A. fumigatus killing was evaluated by the number of CFUs after challenging murine macrophages (Raw 264.7) with conidia from each of the null mutants at 6 h post-infection (hpi), while A. fumigatus invasion was determined by differential staining after infecting human lung epithelial cells (A549) with germinating conidia from all deletion strains at 3 hpi.For statistical analyses, we only considered those strains showing a 30% increase or decrease in conidial survival and invasion compared to the parental strain (A1160).Overall, we identi ed 18 A. fumigatus genes encoding surface proteins that contribute to survival to macrophage killing (9 conidial surface null strains were more susceptible to macrophage killing and another 9 were more resistant to being cleared) (Fig. 3a, Table 1).On the other hand, 27 A. fumigatus genes encoding for surface proteins were important for epithelial cell invasion (3 mutants were more e ciently taken up by the epithelial cells whilst 24 exhibited defects in cell invasion) (Fig. 3b, Table 1).Although there is no correlation between macrophage survival and epithelial cell invasion (Fig. 3c), interestingly, 14 A. fumigatus surfome proteins were identi ed to be involved in both killing and invasion (ΔAFUB_000410, ΔAFUB_004010, ΔAFUB_012100, ΔAFUB_014430, ΔAFUB_028320, ΔAFUB_040910, ΔAFUB_045200, ΔAFUB_047280, ΔAFUB_056750, ΔAFUB_059140, ΔAFUB_059210, ΔAFUB_069520, ΔAFUB_078110 and ΔAFUB_094680) (Figs. 3a and 3b, Table 1), half of them with unknown function, suggesting that a subset of A. fumigatus conidial surfome proteins identi ed in our proteomic analysis contributes to triggering immune cell response.The capacity of the null surface protein strains to induce epithelial cell damage was also evaluated with 3 strains producing less damage than the A1160 strain and only one triggering higher cell toxicity (Fig. 3d, Table 1).
Our data in Fig. 1e to h indicate A. fumigatus induces a stronger immune response in BMDMs compared to three other closely related species.To investigate whether increased induction of immune response in A. fumigatus is driven by speci c conidial surface protein(s), we investigated if the absence of these 42 proteins could play a role in cytokine response by determining the IL1-β and TNF-α production in BMDMs (Figs. 3e and 3f and Table 1).We observed a signi cantly increased 1L-1β production in 11 mutants (AFUB_000410, AFUB_028320, AFUB_031630, AFUB_035440, AFUB_036380, AFUB_039800, AFUB_059140, AFUB_059210, AFUB_068800, AFUB_078110, and AFUB_094680) and decreased production in 3 mutants (AFUB_020170, AFUB_045200, and AFUB_050560) (Fig. 3f and Table 1).TNF-α production was signi cantly increased in AFUB_012100 and decreased in AFUB_010730, AFUB_039800, and AFUB_059210 (Fig. 3f and Table 1).We also evaluated the BMDMs survival in the presence of these mutants by looking at their lactate dehydrogenase (LDH) activity during infection with the mutants compared to the wild-type strain; only the AFU_065340 (a putative zinc-containing alcohol dehydrogenase) mutant had about 80% inhibited LDH.
Using information about A. fumigatus regulatory networks publicly available at the National Center for Biotechnology Information (NCBI) and the clustering online tool FungiExpressZ (https://cparsania.shinyapps.io/FungiExpresZ/),we identi ed two clusters enriched for RNAseq datasets of A. fumigatus exposed to animal cells or invasive aspergillosis conditions (Fig. 4a, highlighted in red circles).These regulatory datasets were obtained after challenging human dendritic and lung epithelial cells with A. fumigatus strains (NCBI Bioprojects PRJEB1583 and PRJNA399754).These two clusters contain 22 genes, 11 of which are also among the 62 A. fumigatus-speci c genes, suggesting that the protein products of these 11 surfome genes could be important for host-pathogen interactions and the establishment of the infection (Figs.4b and 4c, Table 1).Other upregulated clusters are mainly related to response to temperature changes, oxidative and osmotic stresses, or nutritional availability, thus illustrating many other functions of A. fumigatus conidial surface proteins (Supplementary Figure S3; Supplementary Table S3).Furthermore, 7 of the 62 A. fumigatus-speci c genes were within the 1,700 genes whose evolutionary rate differed between pathogenic and non-pathogenic species from Aspergillus section Fumigati 21 , suggesting that they may be associated with the repeated evolution of pathogenicity in Aspergillus spp.(Fig. 4d).
These results suggest that some of the proteins speci cally identi ed in the A. fumigatus surfome, but which are absent from three other closely related species, are important for mediating host interactions and eliciting cytokines, and are modulated at transcriptional level in the presence of animal cells.
Characterization of the glycosylasparaginase null mutant.As an initial step to investigate in more detail the surfome mutants, we prioritize the mutants that elicit increased IL-1β production (Fig. 3g and Table 1) and decided to start by the characterization of ΔAFUB_059140 mutant (here named ΔaspA).The aspA null mutant showed increased killing by BMDMs compared to the wild-type strain but decreased invasion of A549 alveolar epithelial cells (Figs. 3a to 3c and Table 1).We constructed a second independent ΔaspA mutant and both independent aspA null mutants have the same phenotypes, strongly suggesting that the observed phenotypes are only due to the aspA single deletion and not to secondary mutations present in these strains.The aspA gene encodes a putative glycosylasparaginase (Fig. 4e) that catalyzes the hydrolysis of N4-(beta-N-acetyl-D-glucosaminyl)-L-asparagine yielding as products N-acetyl-betaglucosaminylamine plus L-aspartate cleaving the GlcNAc-Asn bond that links oligosaccharides to asparagine in N-linked glycoproteins, playing a major role in the degradation of glycoproteins (http://pfam-legacy.xfam.org/;PF01112).AspA is distributed among several fungal classes (687 sequences) but enriched in Eurotiomycetes and Sordariomycetes where most of the plant and animal pathogens are present (Fig. 4f).The ΔaspA mutants are more phagocytosed by BMDMs than the wildtype (Fig. 5a) and have increased percentage of adherent conidia on the BMDMs surface, and decreased cell viability as measured by CFW uorescence and metabolic activity by MTT, respectively (Figs. 5a to  5c).The increased expression of 1L-1β in the ΔaspA mutants is dependent on their viability since UVkilled conidia have no IL-1β and TNF-α activities comparable to the wild-type strain (Fig. 5d).Although BMDMs have comparable Reactive Oxygen Species (ROS) production upon 4 and 24 h exposure to either the wild-type or ΔaspA mutant, BMDMs have signi cant increased production of ROS when 8 h in the presence of ΔaspA mutant, suggesting that this mutant is either not able to cope with ROS detoxi cation at this time-point or induces increased ROS production (Fig. 5e).The lack of aspA does not change A. fumigatus ability to evade phagolysosome acidi cation in contrast to the A. fumigatus ΔpksP mutant (Fig. 5f).Transwell migration assays show that most of the increased production of IL-1β by aspA null mutants need contact with BMDMs (Fig. 5g).In contrast, TNF-α production is equally dependent on contact for either the wild-type and ΔaspA mutants (Fig. 5h).Taken together, these results suggest that AspA activity is important for BMDMs phagocytosis and viability of A. fumigatus conidia, and that AspA can modulate the IL-1β production by direct contact of the conidia with BMDMs.
AspA is important for establishment of virulence in an immunocompetent murine model of aspergillosis. A. fumigatus ΔaspA have the same virulence than the wild-type in a chemotherapeutic murine model but they are less virulent than the wild-type strain in an immunocompetent murine model as measured by fungal burden (Figs.6a and 6b).We measured the production of cytokines TNF-α, IL-1β, IL-18, IL-12, INF-ϒ, IL-6, and the chemokine CXCL-1 in the lungs homogenates of the immunocompetent model infected by the wild-type and ΔaspA mutants (Figs. 6c to 6i).The lack of AspA caused increased production of proin ammatory cytokines and CXCL-1 that acts as a chemoattractant for several immune cells, and it has already been observed as induced upon A. fumigatus lung infection 22 .Our results indicate that AspA is important for attenuating the in ammatory responses and decreasing the neutrophil migration during the early steps of A. fumigatus infection in the lungs.
Characterization of the heterologously expressed glycosylasparaginase. We heterologously expressed AspA in Pichia pastoris and veri ed if the enzyme was active by checking its putative asparaginase activity 23 (Supplementary Figure S4).Incubation of BMDMs with increasing AspA concentrations had no signi cant impact on LDH activity (Fig. 7a) AspA 0.2 or 2 µg decreased BMDMs viability by about 1 and 10%, respectively, while DMSO decreased by 80% (Fig. 7b).AspA can modulate IL-1β and TNF-α production by using different AspA concentrations in a linear dose-response relationship (Figs.7c and   7d).The cytokine induction by AspA is not related to its glycosylation since AspA exposure to Peptide:Nglycosidase F (PNGase F), which cleaves between the innermost GlcNAc and asparagine residues of high mannose, hybrid, and complex oligosaccharides from N-linked glycoproteins and glycopeptides 24 , did not abolish this induction (Supplementary Figure S5).We were unable to identify asparaginase activity in AspA, and we decided to inactivate the enzyme by boiling it for 5 minutes.Boiling of the AspA decreased IL-1β production by about 50% (Fig. 7e).Most surprisingly, AspA was also able to induce TNF-α production and AspA inactivation decreased its accumulation by 30% (Fig. 7f).Next, we investigated if previous (by 8 h) or concomitant exposure of BMDMs to 2µg of AspA and conidia could change the conidial viability (Fig. 7g).Both treatments increased conidial viability by about 20%, strongly indicating that AspA contributes to conidial viability (Fig. 7g).
Taken together these results suggest AspA modulates BMDM cytokines and it is important to increase conidial viability.
Proteomic pro ling of BMDMs exposed to AspA.To begin investigating which macrophage metabolic pathways are modulated by AspA, we incubated BMDMs in the presence of AspA (0.2 µg) for 24 h, extracted the proteins, and identi ed them by mass spectrometry (Supplementary Table S4).Principal Component Analysis (PCA) and quantitative analysis of these samples showed that BMDM proteins and BMDM proteins exposed to AspA displayed very different distributions (Fig. 8a).Differentially expressed proteins are de ned as those with a minimum of two-fold change in protein abundance (log2FC ≥ 1.0 and ≤ -1.0; FDR of 0.05) when compared to the BMDMs under the equivalent conditions.We observed 101 proteins upregulated and 259 downregulated in the BMDMs exposed to AspA (Fig. 8b and Supplementary Table S4).
We have used the Reference Database of Immune cells (http://refdic.rcai.riken.jp/document.cgi)for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of distinct biological functions of the shared and speci c proteomic differences.The BMDM proteins downregulated in the presence of AspA are involved in (i) cell redox homeostasis, such as glutathione, metabolic process, oxidation reduction, and response to oxidative stress; (ii) carbohydrate metabolism, such as pentose-phosphate shunt, tricarboxylic acid cycle, and glycolysis/gluconeogenesis; (iii) protein synthesis; (iv) pyruvate metabolism and (v) ubiquitin-dependent protein catabolism (Figs.8c and 8d).In contrast, protein transport, vesicle-mediated transport, mRNA processing, and metabolic processes are upregulated (Fig. 8c).We have also visually detected the downregulation of 18 proteins speci cally related to the modulation of the innate immune and lysosomal functions (Table 2and Supplementary Table S4), including: (i) Csf1r, macrophage colony-stimulating factor 1 receptor, a tyrosine-protein kinase that acts as a cell surface receptor and plays a role in the regulation of survival, proliferation and differentiation of mononuclear phagocytes, such as macrophages and monocytes (www.uniprot.org/uniprotkb/P09581/entry);(ii) Ncf1, neutrophil cytosol factor 1, required for the activation of the latent NADPH oxidase (www.uniprot.org/uniprotkb/Q09014/entry);(iii) Cotl1, coactosinlike protein that in uences leukotrienes synthesis (www.uniprot.org/uniprotkb/Q9CQI6/entry);(iv) Mtdh, protein LYRIC that activates the nuclear factor kappa-B (NF-kappa B) transcription factor (www.uniprot.org/uniprotkb/Q80WJ7/entry),and (v) and a few lysosomal proteases cathepsins; and upregulated, (i) Ptprc, a receptor-type tyrosine-protein phosphatase C that is required for T-cell activation through the antigen receptor (www.uniprot.org/uniprotkb/P06800/entry)and (ii) Ctss, cathepsin S, a key protease responsible for the removal of the invariant chain from MHC class II molecules and MHC class II antigen presentation (www.uniprot.org/uniprotkb/O70370/entry).Examination of the general mouse functional protein association network, retrieved from STRING (https://string-db.org)showed that 15 of these proteins have functional associations, suggesting that AspA impacts speci c protein interaction networks (Fig. 8e).
Taken together, our results suggest that AspA modulates several BMDM metabolic pathways speci cally affecting the abundance of proteins important for immune function.

DISCUSSION
A. fumigatus is a prominent human fungal pathogen that infects and frequently kills hundreds of thousands yearly.Most other closely related species, on the other hand, are not pathogenic 2,15,25 .The dozen or so clinically relevant Aspergillus species are scattered among hundreds of non-pathogenic species across the Aspergillus phylogeny, implying that the capacity to cause illness or pathogenicity has evolved multiple times in this lineage 2 .The observed pathogenicity spectrum cannot be solely explained by changes in species ecologies or ascertainment bias, suggesting that the repeated evolution of Aspergillus pathogenicity has a genetic foundation, at least in part 2 .A. fumigatus conidia are the rst fungal structure that encounters the host.In Immunocompetent hosts, conidia are cleared and do not affect host health.However, there are several mechanisms of conidial and germling recognition and evasion based mostly on melanin and polysaccharides, such as β-1,3 glucan, chitin, and galactosaminogalactan present on the conidial surface.Although the importance of protein effectors for pathogenesis and virulence have been extensively described in plant fungal pathogens, there are very few reports about conidial surface proteins or protein effectors in human fungal pathogens.In A. fumigatus, these include hydrophobin 6 , the conidial surface protein cpcA 8 that prevents immune recognition of β-1,3-glucans, and the HscA protein that anchors human p11 on phagosomal membranes, rewiring the vesicular tra cking to the non-degradative pathway, allowing escape of conidia 9 .More recently, a Cryptococcus neoformans secreted effector that triggers allergic in ammation 26 was also identi ed.
Here, we used a phylogenetic approach to investigate proteins speci cally expressed in A. fumigatus conidial surface by comparing A. fumigatus proteins with two closely non-pathogenic related species, A and Aoe, and one more distantly-related pathogenic species, Ale.These surface proteins could be potential surface proteins or protein effectors that modulate different aspects of virulence and pathogenicity or alternatively new genetic determinants important for maintaining the correct conidial structure allowing attachment to host surfaces or preventing immune recognition.Using this approach, we identi ed 62 A. fumigatus-speci c conidial surface proteins and were able to delete 42 of them.Some of these proteins affect the correct assembly of the conidial surface components since they have increased or decreased chitin exposure or different hydrophobic properties.However, several of them affect the interaction with the host cells since they modulate survival in the presence of macrophages, invasion and/or damage to epithelial cells, and cytokine production.This collection of genes and mutants provides an opportunity for the characterization of genetic determinants important for the A. fumigatus-host interaction.
We focused on one of these genetic determinants, a gene encoding a putative glycosylasparaginase, named AspA.To our knowledge this is the rst characterized fungal glycosylasparaginase. Homologues of this gene are not present in Saccharomyces cerevisiae and when mutated in humans (AGA gene encodes the glycosylasparaginase) cause the most frequent type of recessively inherited lysosomal storage disease named Aspartylglucosaminuria (AGU) 27 .AGU is a disorder related to degradation of glycoproteins.Carbohydrate moieties are typically found in glycoproteins and are linked to the protein moiety by an N-glycosidic bond formed by the amino acid L-asparagine and the monosaccharide Nacetylglucosamine.The lysosomal resident enzyme glycosylasparaginase, which cleaves the N-glycosidic link between the L-asparagine and N-acetylglucosamine moieties of GlcNAc-Asn, is lacking in AGU.This enzyme de ciency causes a buildup of undegraded aspartylglucosamine and several other glycoasparagines, i.e., glycoconjugates having a L-asparagine moiety attached to the carbohydrate chain, in the affected person's tissues and body uids 28,29 .AGU patients have developmental delays, hyperactivity, early growth spurt, inguinal and abdominal hernias, clumsiness, characteristic facial features, recurrent respiratory and ear infections, and chronic arthritis 30 .
As far as we know there is no previous information about the involvement of glycosylasparaginases in bacterial and fungal pathogenesis.However, there are several reports about possible roles played by asparaginases in bacterial virulence [31][32][33][34][35] , but not in fungal virulence.Several bacterial virulence factors have evolved to use enzymatic deamidation, i.e., the irreversible conversion of the amino acids glutamine and asparagine to glutamic acid and aspartic acid, respectively 32 .Helicobacter pylori cytotoxic activity was signi cantly decreased by an asparaginase-de cient strain 31 while Mycobacterium tuberculosis exploits asparaginase to assimilate nitrogen to produce ammonium and resist acid stress during infection 33 ; Salmonellae asparaginases mediate virulence and inhibit T cell responses 34,35 .A. fumigatus has two putative glycosylasparaginase (AFUB_059140, which is described here, and AFUB_021340, none of them with a putative signal peptide) and two asparaginase encoding-genes (AFUB_003170 and AFUB_051450, with and without a putative signal peptide, respectively).It remains to be investigated if these genes are also important for A. fumigatus virulence.
The ΔaspA mutants are more phagocytosed and killed than the wild-type, elicit more ROS production and need BMDMs contact to have increased IL-1β production, but have identical percentages of acidi cation of the phagolysosomes.These results suggest that AspA has a role in earlier or later steps after the formation of phagolysosomes.AspA is important for the establishment of the infection in an immunocompetent mouse model but not in the chemotherapeutic mouse model, and the lack of AspA also plays an important role in modulating proin ammatory cytokines in vivo.The lack of AspA causes increased production of proin ammatory cytokines and CXCL-1 that acts as a chemoattractant for several immune cells, especially neutrophils, and it has already been observed as induced upon A. fumigatus lung infection 22 .Our results indicate that AspA is important for modulating the in ammatory response and neutrophils migration upon A. fumigatus infection in the lungs.
We also investigated AspA through its heterologous expression and the in uence of AspA on cytokine production and macrophage metabolism.Although the human glycosylasparaginase has both deglycosylation and asparaginase activities 36 , we were not able to detect asparaginase activity in A. fumigatus AspA.However, inactivated boiled AspA decreases BMDM cytokine induction, suggesting that it is still active, and this activity is not due to possible glycosylated residues present in AspA since PNGase treatment has not abolished the cytokine induction.AspA has no putative signal peptide, was identi ed only at the surface of the dormant conidia and does not seem to be secreted.When BMDMs were exposed to AspA, there was a dramatic change in its proteomic pro ling, with decreased cell redox and carbohydrate, and pyruvate metabolism, and protein synthesis suggesting that AspA is affecting the BMDMs aerobic metabolism and ROS production, and decreasing the expression of proteins important for ROS production, such as Ncf1, Neutrophil cytosol factor 1, a cytosolic subunit of neutrophil NADPH oxidase, a multicomponent enzyme that is activated to produce superoxide anion.It has been previously shown that patients with hypersensitivity pneumonitis (HP), a rare initial presentation in chronic granulomatous disease (CGD) have mutations in NCF1 and NCF2 genes and are more susceptible to invasive pulmonary A. fumigatus infection 37 .Actually, the ΔaspA shows higher induction of ROS in BMDMs.We also observed that BMDMs exposure to AspA decreased the production of several lysosomal proteases, cathepsins, important for protein degradation.
Upon conidial phagocytosis, phagosomes are fused with lysosomes resulting in functional phagolysosomes that are integral part of a degradative pathway that kills and destroys conidia.Is AspA important for escaping phagolysosomal killing?An attractive model for AspA mechanism of action could be that upon conidial phagocytosis, AspA is already interacting with proteins at the macrophage surface deglycosylating asparagine residues and modifying protein activity.Since glycosylasparaginases are described as lysosomal resident proteins, they must be resistant to acid pH (AspA has an isoelectric point of 5.4).Upon conidial phagocytosis and during the formation of phagolysosomes, AspA remains active in the phagolysosome, eventually deglycosylating asparagine residues affecting conidial survival.This apparently has no effect on the percentage of acidi cation of the phagolysosome since there are no differences between the mutant and wild-type strains.However, it is possible after partial conidial destruction, AspA remains active in the host lysosome affecting the glycosylation metabolism in the cell, and consequently allowing some conidia to survive. A. fumigatus conidia can survive for some time in the phagosomes of immune cells and epithelial cells [38][39][40][41] .AspA could affect host's membrane tra cking system or redirect the conidia-containing phagosomes toward exocytosis and release of conidia.This has already been reported for a novel A. fumigatus surface effector, the heat-shock protein HscA that interacts with the human p11 protein, a decisive regulatory node for directing endosomes to different pathways 9 .Different aspects of our proposed model remain to be investigated.

METHODS
Fungal strains.All strains included in this work are listed in Table 1.Knockout mutant strains belong to the Manchester Infection Fungal Group (MFIG) collection and were constructed as described in 20 .
Surface proteome analysis.Freshly harvested conidia of A. fumigatus A1163, A. oerlinghausenensis CBS 139183 T , A. scheri NRRL 181, and A. lentulus CNM-CM6069 were cultivated on potato dextrose agar at 37°C for 0 h (resting conidia) or 4 h (swollen conidia).1 × 10 9 conidia of each species in three biological replicates were washed twice with 25 mM ammonium bicarbonate (AB) and centrifuged at 1,800 × g for 10 min.Conidia were resuspended in 800 µl of 25 mM AB and incubated with 5 µg trypsin (MS approved, Serva) for 5 min at 37°C.Cell suspensions were immediately passed through 0.2 µm syringe lters (cellulose acetate, Sartorius) to separate conidia from the digestion buffer followed by washing the lters with 200 µl of 25 mM AB. Subsequently, 10 µl of 90% (v/v) formic acid was added to the cell-free solution to stop the proteolytic digestion.Samples were evaporated to dryness in a vacuum concentrator (Eppendorf), resuspended in 30 µl of 2% (v/v) acetonitrile (ACN) and 0.05%(v/v) tri uoroacetic acid (TFA), centrifuged for 15 min at 14,000 × g through 10 kDa molecular weight cut-off lters (modi ed PES, VWR), and transferred into HPLC vials.LC − MS/MS analysis was performed on an Ultimate 3000 RSLC nano instrument coupled to a QExactive HF mass spectrometer (Thermo Fisher Scienti c) as described previously 12  ions were scanned at m/z 300-1500, a mass resolution of 60,000 full width at half maximum (FWHM), an automatic gain control (AGC) target of 1 × 10 6 , and a maximum injection time (maxIT) of 100 ms.
Precursor ions were isolated with a width of m/z 2.0.Fragment ions generated in the higher-energy collisional dissociation (HCD) cell at 30% normalized collision energy using N 2 gas were scanned at 15,000 FWHM, an AGC target of 2×10 5 , and a maxIT of 80 ms.Dynamic exclusion was set to 25 s.
The MS/MS data were searched against the FungiDB databases of A. fumigatus Af293 and A. scheri NRRL_181, the NCBI database of A. lentulus and the JGI database of A. oerlinghausenensis (all downloaded on 2020/10/27) using Proteome Discoverer 2.4 and the algorithms of Mascot 2.4.1,Sequest HT, and MS Amanda 2.0 and MS Fragger 3.0.Two missed cleavages were allowed for the tryptic digestion.The precursor mass tolerance was set to 10 ppm and the fragment mass tolerance was set to 0.02 Da.Modi cations were de ned as dynamic Met oxidation, protein N-term acetylation and/or loss of methionine.A strict false discovery rate (FDR) < 1% (peptide and protein level) and a search engine score of > 30 (Mascot), > 4 (Sequest HT), > 300 (MS Amanda) or > 8 (MS Fragger) were required for positive protein hits.The Percolator node of PD2.4 and a reverse decoy database was used for qvalue validation of spectral matches.Only rank 1 proteins and peptides of the top scored proteins were counted (Supplementary Table S1).
The infected macrophage proteome was performed through lysis in 8M urea in 25 mM AB containing protease inhibitor cocktail using a probe tip sornicator: 40% amplitude for 3 cycles for 10 s and intervals of 10 s.After sonication, samples were centrifuged, the supernatant was transferred to a new tube and quanti ed using Bradford reagent.Proteins were reduced with 10mM DTT for 30 min at 30 degrees, alkylated with 40mM iodoacetamide for 40 min at room temperature in the dark and overnight digested with trypsin in the ratio 1:50 (enzyme to protein ratio).The tryptic peptides were desalted with Oasis HLB Cartridges (Waters) according to the manufacturer instructions, dried down by speed-vac, and then resuspended in 0.1% formic.LC-MS/MS analysis was performed in an EASY-nLC system (Thermo Scienti c) coupled to LTQ-Orbitrap Velos mass spectrometer (Thermo Scienti c).Peptides were separated on C18 PicoFrit column (C18 PepMap, 75 µm id × 10 cm, 3.5 µm particle size, 100 Å pore size; New Objective) using a gradient of A and B buffers (buffer A: 0.1% formic acid; Buffer B: 95% CAN, 0.1% formic acid) at a ow rate of 300nL/min: from 2-30% B over 80 min and from 30-90% B over 5 min.The LTQ-Orbitrap Velos was operated in positive ion mode with data-dependent acquisition.The full scan was obtained in the Orbitrap with an automatic gain control (AGC) target value of 10e6 ions and a maximum ll time of 500 ms.Each precursor ion scan was acquired at a resolution of 60,000 FWHM in the 400-1500 m/z mass range.Peptide ions were fragmented by CID MS/MS using a normalized collision energy of 35.The 20 most abundant peptides were selected for MS/MS and dynamically excluded during 30 sec.All raw data were accessed in the Xcalibur software (Thermo Scienti c).For protein identi cation, raw data were processed using MaxQuant software version 1.5.3.8.The MS/MS spectra were searched against a protein database composed of Aspergillus fumigatus and Human sequences with the addition of common contaminants with a tolerance level of 4.5 ppm for MS and 0.5 Da for MS/MS.Trypsin was selected as a speci c enzyme with a maximum of two missed cleavages.Carbamidomethylation of cysteine (57.021Da) was set as a xed modi cation, and oxidation of methionine (15.994Da), deamidation NQ (+ 0.984 Da) and protein N-terminal acetylation (42.010Da) were set as variable modi cations.PSMs, peptides and proteins were accepted at FDR less than 1%.
Multivariate statistical analyses were performed on the protein groups using the label-free quanti cation (LFQ) results.The analyses were conducted using the LFQ-Analyst web platform 42 with default parameters, including a p-value cutoff of 0.05 and a Log2 fold change cutoff of 1.
BMDMs protein extraction.The BMDMs protein extraction was performed in 8M urea containing protease inhibitor cocktail using a probe tip sornicator: 40% amplitude for 3 cycles for 10 s and intervals of 10 s.
After sonication, samples were centrifuged, the supernatant was transferred to a new tube and quanti ed using Bradford reagent.Proteins were reduced with 10mM DTT, alkylated with 40mM iodoacetamide and overnight digested with trypsin in the ratio 1:50 (enzyme:protein).The tryptic peptides were desalted with Oasis HLB Cartridges, dried down by speed-vac, and then resuspended in 0.1% formic to the LC-MS/MS analysis that was performed in an EASY-nLC system (Thermo Scienti c) coupled to LTQ-Orbitrap Velos mass spectrometer (Thermo Scienti c).Peptides were separated on C18 PicoFrit column (C18 PepMap, 75 µm id × 10 cm, 3.5 µm particle size, 100 Å pore size; New Objective) using a gradient of A and B buffers (buffer A: 0.1% formic acid; Buffer B: 95% ACN, 0.1% formic acid) at a ow rate of 300nL/min: from 2-30% B over 80 min and from 30-90% B over 5 min.The LTQ-Orbitrap Velos was operated in positive ion mode with data-dependent acquisition.The full scan was obtained in the Orbitrap with an automatic gain control (AGC) target value of 10e6 ions and a maximum ll time of 500 ms.Each precursor ion scan was acquired at a resolution of 60,000 FWHM in the 400-1500 m/z mass range.
Peptide ions were fragmented by CID MS/MS using a normalized collision energy of 35.The 20 most abundant peptides were selected for MS/MS and dynamically excluded during 30 sec.All raw data were accessed in the Xcalibur software (Thermo Scienti c).
For protein identi cation, raw data were processed using MaxQuant software version 1.5.3.8.The MS/MS spectra were searched against a protein database composed of human sequences with the addition of common contaminants with a tolerance level of 4.5 ppm for MS and 0.5 Da for MS/MS.Trypsin was selected as a speci c enzyme with a maximum of two missed cleavages.Carbamidomethylation of cysteine (57.021Da) was set as a xed modi cation, and oxidation of methionine (15.994Da), deamidation NQ (+ 0.984 Da) and protein N-terminal acetylation (42.010Da) were set as variable modi cations.Proteins and peptides were accepted at FDR less than 1%.
Glycosylasparaginase heterologous expression.The synthetic asparaginase gene (AFUB_059140) cloned in the expression vector pPICZαA (Invitrogen) was synthesized by the company Genescript.The vector has optimized codons for the expression of Komagataella pha i and the α factor secretion signal in the N-terminal portion and Zeocin resistance gene.The plasmid obtained was propagated in competent Escherichia coli (DH10β) cells (Thermo Fisher Scienti c, Waltham, USA), resistant to zeocin.The puri ed plasmid was linearized by Anza 24 MssI restriction enzyme (Invitrogen), puri ed again, and transformed into competent K. pha i (KM71H) by electroporation (1.5 kV, 25 mF and 200 Ω) in a 0.2 µm cuvette using an electroporator (MicroPulser Electroporator, Bio-Rad).The recombinant strain was selected, cultivated and induced to express the rAsparaginase protein following the EasySelect Pichia Expression Kit manual (Invitrogen, Waltham, USA).After 120 h of induction with 1% (v/v) methanol, the material was centrifuged (5,000 × g, 45 min at 4°C) and the supernatant was ltered through a 0.22 µm lter.
The ltered material was dialyzed against 50 mM Bicine buffer (pH 8.5) and subjected to anion exchange puri cation.The Mono Q 5/50 GL column (GE Healthcare) was equilibrated with 50 mM Bicine buffer (pH 8.5) and the enzyme eluted with 180 mM NaCl.Protein integrity and sample purity were con rmed by running a 12% SDS-PAGE 43 .Puri ed protein fractions were dialyzed against 1X PBS (pH 7.4) using the Vivaspin 10 kDa centrifuge system (Sartorius).Protein quanti cation was performed in a spectrophotometer at 280 nm and the assays were performed with the enzyme at a concentration of 5.88 µM.
AspA phylogenetic analysis.The visualize the distribution of the glycosylasparaginase gene presence among fungi, the AspA amino acid sequence (AFUB_059140) was used for BLAST search 44 against the NCBI RefSeq database 45 .The 687 sequences found were aligned using MAFFT v7.508 46 for the inference of a maximum likelihood phylogenetic tree using IQ-TREE v1.7 47 with the substitution model JTT + I + G4, determined to be the best by the program.The calculated tree was visualized and edited using iTOL v6 48 .
Macrophage infection, cytokine and LDH determination.BMDMs were cultured as described before and were seeded at a density of 10 6 cells/ml in 24-well plates (Greiner Bio-One, Kremsmünster, Austria).The cells were challenged with the conidia of different strains at a multiplicity of infection of 1:10 and incubated a 37°C with 5% (vol/vol) CO 2 for 24h.BMDMs were also stimulated with different concentrations (mM) of AspA protein (desnaturated or not) by boilling per 10 min 100 degrees Celsius.LPS (standard LPS, E. coli 0111: B4; Sigma-Aldrich, 500 ng/mL) plus Nigericin (tlrl-nig, InvivoGen 5 µM/mL) and medium alone were used respectively as the positive and negative controls.Cell culture supernatants were collected and stored at − 80°C until they were assayed for TNF-α, IL-1 and LDH release using Mouse DuoSet ELISA kits (R&D Systems, Minneapolis, MN, USA and CyQUANT™ lactate dehydrogenase (LDH) Cytotoxicity Assay (Invitrogen), according to the manufacturer's instructions.For cytokine determination, plates were analysed by using a microplate reader (Synergy™ HTX Multi-Mode, BioTek) measuring absorbance at 450 nm.Cytokine concentrations were interpolated from a standard curve and statistical signi cance was determined using an ANOVA (GraphPad Prism 8.0, La Jolla, CA).
The level of LDH was determined by measuring absorbance at 490 and 680nm using a microplate reader (Synergy™ HTX Multi-Mode, BioTek).All assays were performed in triplicate in three independent experiments.
Aspergillus growth condition and Fluorescein isothiocyanate (FIT C) label.Aspergillus strains were cultivated on minimal medium (AMM) agar plates at 37°C for 3 days.Conidia were harvested in sterile water with 0.05% (vol/vol) Tween20.The resulting suspension was ltered through two layers of gauze (Miracloth, Calbiochem).FITC-labelling of conidia was performed with 0.1 mg/ml FITC (Sigma) in 0.1 M Na 2 CO 3 at 37°C for 30 min.Labelled conidia were washed three times with PBS, 0.1% (vol/vol) Tween20.
The conidia concentration was determined using a hemocytometer.
Phagocytosis and adhesion BMDMs were cultivated in Modi ed Eagle Medium (DMEM) supplemented with 10% (vol/vol) heat-inactivated fetal calf serum, 2mM glutamine and penicillin-streptomycin.For infection experiments, macrophages were seeded on glass cover slips in 24 well plates at a density of 5×10 5 cells per well and allowed to grow adherently overnight.Following washing with prewarmed medium, FITC-labelled conidia were added at a multiplicity of infection of 10.
The infection experiment was synchronized for 30 min at 4°C.Unbound conidia were removed by washing with pre-warmed medium and phagocytosis was initiated by shifting the coincubation to 37°C in a humidi ed CO 2 incubator.After 1 h the phagocytosis was stopped by washing with ice-cold PBS.
Labelling of extracellular conidia was performed by incubation with PBS, 0.25 mg/ml calco uor white (Sigma) for 30 min at 4°C to avoid further.The cells were washed twice with PBS and xed with 3.7% (vol/vol) formaldehyde/PBS for 15 min followed by two washes with PBS.Microscopic photographs were taken on a Zeiss microscope.For statistical reproducibility two biological replicates and in each case two technical replicates were made and analyzed for each strain.The phagocytic index was enumerated by counting 100 macrophages per cover slip from duplicate wells.Phagocytic index was calculated by the average number of conidia that had been phagocytosed for each macrophage.
Macrophage killing assay.BMDMs were seeded at a density of 10 6 cells/ml in 24-well plates (Corning Costar) and were challenged with conidia at a multiplicity of infection of 1:10 and incubated a 37°C with 5% (vol/vol) CO for 24h.After media was removed the cells were washed with ice-cold PBS and nally 2ml of sterile water was added to the wells.A P1000 tip was then used to scrape away the cell monolayer and the cell suspension was collected.This suspension was then diluted 1:1000 and 100 µl was plated on Sabouraud agar before the plates were incubated a 37°C overnight and the colonies were counted.50 µl of the inoculum adjusted to 10 3 /ml was also plated on SAB agar to correct CFU counts.
The CFU/ml for each sample was calculated and compared to the A1160 wild-type strain.
Acidi of phagolysosomes.Acidi cation of phagolysosomes was essentially determined as previously described 17 .In brief, RAW 264.7 murine macrophages were incubated in Dulbecco's Modi ed Eagle Medium (DMEM, Gibco) with 27.5 mg/mL gentamicin sulfate (Gibco), 10% (vol/vol) fetal bovine serum (FBS, GE Healthcare Life Sciences), and 1% (wt/vol) ultraglutamine (Gibco) at 37°C and 5% (vol/vol) CO 2 in humidi ed incubator.Cells were pre-stained with 50 nM LysoTracker Red DND-99 (Invitrogen) for 1 h before infection.Conidia were stained with 0.1 mg/mL Calco uor White (CFW, Sigma Aldrich) for 10-15 min before infection.1 x 10 5 cells per well were added to 8 well ibidi slides (ibidi GmbH) and infected with an MOI of 3 with CFW stained conidia for 2 h at 37°C and 5% (vol/vol) CO 2 in a humidi ed incubator.Uptake of conidia was synchronized by centrifugation at 100 x g for 5 min.After infection, cells were washed with PBS and xed using 3.7% formaldehyde in PBS for 10 min, followed by a washing step with PBS.Slides were directly imaged using a Zeiss LSM 780 microscope and analyzed using Zeiss ZEN software.At least, 100 conidia were counted for each biological replicate.Clear red signals surrounding conidia within phagosomes were evaluated as acidi ed conidia.Three independent experiments were performed.
XTT assay was performed as described 50,51 .Brie y, 5 × 10 5 BMDMs were plated in 96-well and then infected with 5 × 10 6 spores in a nal volume of 100 µl DMEM media.Control wells contained spores but no BMDMs or only BMDMs.Following 2 hr incubation, the macrophages were subjected to hypotonic lysis by three gentle washes with distilled water followed by a 30 min incubation with distilled water at 37°C.Supernatants then were removed, with great care taken not to remove the spores.DMEM media without Phenol Red containing 400 µg/ml of XTT and 50 µg/ml of Coenzyme Q, were added, and the wells were incubated for 2 hr at 37°C.The OD 450 and OD 650 were then measured, and data were expressed as the of antifungal activity according to the published formula 50 .
Cytokine and chemokine quanti cation.For lung homogenates, the lungs of all experimental groups were homogenized in PBS supplemented with Complete Mini protease inhibitor tablets (Roche), clari ed by centrifugation, and stored at − 80°C.A panel of cytokines and chemokines were quanti ed by ELISA (R&D Systems) according to the manufacturer's instructions.
Detection of reactive oxygen species (ROS).In 96-well tissue culture (TC)-treated dark clear bottom plates (Sigma-Aldrich), 5×105 BMDMs per 100 µL were infected with 5×106 conidia followed by 4, 8, and 24 h of incubation at 37 0 C. The cells were treated with the ROS detection dye CellROX ® Reagent (ThermoFisher) at a nal concentration of 5 µM for 30 min at 37 °C.ROS detection was performed using a uorescent plate reader (Thermo Fisher), with an excitation of 485 nm and an emission of 520 nm.BMDMs stimulated with 50 nM phorbol-12-myristate-13-acetate (PMA) (Sigma-Aldrich) for 20 min at 37 °C were used as positive control.
Transwell Assay.BMDMs were seeded at a density of 10 6 cells/ml in 24-well plates (Corning Costar) and the assay was performed as the addition each well containing a transwell insert (0.2-µm lter; Corning) containing medium alone, ΔaspA-1, ΔaspA-2 or wild-type conidia at a multiplicity of infection of 1:10 and incubated a 37°C with 5% (vol/vol) CO 2 for 24h.After incubation cell culture supernatants were collected and stored at − 80°C until they were assayed for TNF-α, IL-1 production.In the transwell system the small and soluble compounds are able to migrate between the upper transwell and lower chamber, whereas spores are prevented from moving between the chambers.
Epithelial cell invasion assay.The capacity of the various strains to invade the A549 cells was determined using previously described methods with some modi cation 52,53 .The A. fumigatus wild-type and mutant strains were grown on Sabouraud dextrose agar (Difco) at 37°C for 7 d prior to use.Conidia were harvested with phosphate-buffered saline (PBS) containing 0.1% (vol/vol) Tween 80 (Sigma-Aldrich) and enumerated with a hemacytometer.The A549 pulmonary epithelial cell line (American Type Culture Collection) was cultured in F12 medium (American Type Culture Collection) containing 10% (vol/vol) fetal bovine serum (Gemini Bio-Products), and 2 mM L-glutamine with penicillin and streptomycin (Irvine Scienti c) in 5% CO 2 at 37°C.Before the assay, 2 x 10 5 A549 cells were cultured in 24-well tissue culture plates containing bronectin coated circular glass coverslips in each well for overnight.A. fumigatus conidia were pre-germinated in Sabouraud dextrose broth (Difco) at 37°C for 5.5 h, counted and suspended in F12k medium.Next, 10 5 germlings of each strain in 1 ml F12 K medium were added to A549 cells that had been grown to con uency on the glass coverslips.After incubation for 3 h, the cells were rinsed with 1 ml HBSS in a standardized manner and then xed with 4% (vol/vol) paraformaldehyde.The noninternalized portions of the organisms were stained with a polyclonal rabbit anti-A.fumigatus primary antibody (Meridian Life Science, Inc.) followed by an AlexaFluor 568-labeled secondary antibody (Life Technologies).After the coverslips were mounted inverted on microscope slides, they were viewed by epi uorescence.The number of cell-associated organisms was determined by counting the number of red uoresced organisms per eld (HPF).The number of endocytosed organisms was determined by subtracting the number of non-internalized organisms (staining of entire germlings) from the number of cell-associated organisms.At least 100 organisms per coverslip were scored and each strain was tested in triplicate.
Epithelial cell damage assay.To evaluate the capacity of various strains to damage A549 epithelial cells, we used our standard 51 Cr release assay [53][54][55] .A. fumigatus conidia and A549 were prepared as described before.The A549 cells were grown to con uency in a 24-well tissue culture plate and then loaded with 51 Cr overnight.The following day, the cells were rinsed twice with HBSS to remove the unincorporated 51 Cr and 5 x 10 5 conidia in 1 ml of F12K medium of each strain were added to triplicate wells.After 16 h of incubation, 500 µl of medium above the cells were collected and transferred to glass test tube A. Next, the remaining medium in the wells was collected and placed in glass test tube B. After lysing the A549 cells with 6 N NaOH, the lysate was collected and the wells were rinsed twice with RadiacWash (Biodex Medical Systems).The lysate and rinses were added to test tube B. The amount of 51 Cr in the medium and the cell lysate was measured using a gamma counter.The spontaneous release of 51 Cr was determined using uninfected A549 cells that were processed in parallel.The speci c release of 51 Cr was calculated using our previous described formula.Each experiment was performed in triplicate.Mutants that caused less damage to the A549 cells than the control strain were tested at least one more time to verify the results.
Animal survival curves and burden.Inbred female mice (BALB/c or C57BL/6 strains; body weight, 20-22 g) were housed in vented cages containing ve animals.Mice were immunosuppressed with cyclophosphamide (150 mg/kg of body weight), which was administered intraperitoneally on days − 4, -1 and 2 prior to and post infection (infection day is "day 0").Hydrocortisonacetate (200 mg/kg body weight) was injected subcutaneously on day − 3. Mice (10 mice per group) were anesthetized by halothane inhalation and infected by intranasal instillation of 20 µL of 1.0 x 10 5 conidia of A. fumigatus wild-type or mutant strains, A. lentulus, A. oerlinghausenensis, or A. scheri (the viability of the administered inoculum was determined by incubating a serial dilution of the conidia on MM medium, at 37°C).As a negative control, a group of 10 mice received PBS only.Animals were sacri ced 15 days postinfection.To investigate fungal burden in murine lungs, mice are immunosuppressed as described previously or not, and mice were intranasally inoculated with 1 x 10 6 conidia/20 µl of suspension for the chemotherapeutic murine model and with 5 x 10 8 conidia/20 µl for the immunocompetent murine model.
Animals were sacri ced 48 h post-infection, and the lungs were harvested and immediately frozen in liquid nitrogen.DNA was extracted via the phenol/chloroform method and 400 mg of total DNA from each sample were used for quantitative PCRs using primers to amplify the 18S rRNA region of A. fumigatus and an intronic region of mouse GAPDH (glyceraldehyde-3-phosphate dehydrogenase).Six-point standard curves were calculated using serial dilutions of gDNA from A. fumigatus strain and the uninfected mouse lung.Fungal and mouse DNA quantities were obtained from the threshold cycle (CT) values from the appropriate standard curves.The qPCR analysis was performed using the SYBR green PCR master mix kit (Applied Biosystems) in the ABI 7500 Fast real-time PCR system (Applied Biosystems, Foster City, CA, USA).
The principles that guide our studies are based on the Declaration of Animal Rights rati ed by UNESCO on January 27, 1978  of Fc-h-dectin-hFc (Invivogen) was added to the UV-irradiated conidia and incubated for 1 h at RT, followed by the addition of 1:1,000 DyLight 594-conjugated goat anti-human IgG1 (Abcam) for 1 h at RT. Conidia were then washed with phosphate-buffered saline (PBS), and uorescence was read at 587-nm excitation and 615-nm emission.For chitin staining, 200µl of a PBS solution with 10mg/ml of calco uor white (CFW) was added to the UV-irradiated conidia, which were incubated for 5 min at RT and washed with PBS before uorescence was read at 380-nm excitation and 450-nm emission.For N-acetyl-Dglucosamine (GlcNAc) staining, 200 µl of PBS supplemented with 0.1 mg/ml of wheat germ agglutinin (WGA) lectin (lectin-FITC L4895; Sigma) was added to the UV-irradiated germlings for 1 h at RT.
Germlings were washed with PBS, and uorescence was read at 492 nm excitation and 517 nm emission.
All experiments were performed using at least 4 repetitions, and uorescence was read in a microtiter plate reader (Synergy HTX Multimode Reader; Agilent Biotek or EnSpire Multimode Plate Reader; Perkin Elmer).
Fungal adhesion assay.In order to determine the adhesion capacity of conidia from deleted mutant and A1160 strains, 10 4 conidia were inoculated into 200 ul of MM liquid medium in a 96-well polystyrene microtiter plate.Following an initial adherence phase of 4 h during static incubation at 37°C (swollen conidia) or 4°C (resting conidia), unbound conidia were washed with sterile PBS.Fresh MM liquid medium was added to the adhered conidia, and static submerged cultures were grown for up to 24 h at 37°C.Subsequently, the plate was washed exhaustively with PBS prior to incubation with 200 ul 0.5% (wt/vol) crystal violet solution for 5 min at room temperature.The stained mycelia were then exhaustively washed with sterile water and air dried.Finally, the crystal violet was eluted from the wells using 100% ethanol, and the absorbance was measured at 590 nm in a Synergy HTX Multimode Reader (Agilent Biotek).
Assessment of conidial hydrophobic properties.The distribution of the conidia on a water-oil interface was performed by the addition of 1x10 8 conidia of the wild-type and each deleted strain in a solution containing water and tributyrin (1:1 [vol/vol]).The mixture was vortexed during 1 min and set to allow the conidia to disperse for 1 h.The distribution of the conidia for each mutant were then visually compared to the A1160 wild-type strain.
Statistical analysis.signi cance for each experiment was established by one-way or two-way ANOVA using the built-in tools of Prism 6. Statistical tests are indicated in the gure legends and were chosen based on the nature of the experiment and the standard tests employed in the eld.Underlying assumptions for these tests, including sample independence, variance equality, and normality were assumed to be met.One-way ANOVA was followed by Tukey's or Dunnett's multiple comparison test when appropriate.Asterisks denote p value as follows: ns = not signi cant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Declarations Data Availability
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository 57 with the dataset identi ers PXD031199 (Username: reviewer_pxd031199@ebi.ac.uk;Password: HjEpfE78) and PXD044190 (Username: reviewer_pxd044190@ebi.ac.uk;Password: WNvL2PfA).

Tables
Tables 1 and 2 are available in the Supplementary Files section.
Figures   2023).e.Growth phenotypes of the wild-type A1160 strain (in red) and deleted mutants grown for 72 h at 37˚C or 44˚C in solid MM media and MM supplemented with Congo Red (CR; 10µg/ml) or hydrogen peroxide (H 2 O 2 ; 1.5mM).Red circle represents A1163 strain.f.Germination rates of the wild-type A1160 strain (in blue) and deleted mutants.Only those hits below or above the 30%-fold change threshold and stastistically different from the parental strain A1160 are displayed in the gure.Blue circle represents A1163 strain.g.Adhesion, masured by CV assay, of resting (○) and swollen (△) conidia for all all mutant strains and A1160 (in blue).Only hits that shared the same signi cant phenotype (lower and higher detection in green and red, respectively) in both stages are highlighted in the gure.Red circle and triangle represent A1163 strain.h.Detection of chitin (CFW), Nacetylglucosamine (GlcNAc) (WGA) and β-(1,3)-glucan (Dectin) contents on the conidial surface of all mutant strains and A1160 (in blue) in resting (○) and swollen (△) conidia.Blue circle and triangle represent A1163 strain.Only hits that shared the same signi cant phenotype (lower and higher detection in green and red, respectively) in both stages are highlighted in the gure.

Figure 1 A
Figure 1

Figure 3 A
Figure 3

Figure 4 Some
Figure 4

Figure 5 A
Figure 5

Figure 6 A
Figure 6

Figure 7 Heterologously
Figure 7 in its 8th and 14th articles.All protocols adopted in this study were approved by the local ethics committee for animal experiments from the University of São Paulo, Campus of Ribeirão Preto (Permit Number: 08.1.1277.53.6;Studies on the interaction of A. fumigatus with animals).Groups of ve animals were housed in individually ventilated cages and were cared for in strict accordance with the principles outlined by the Brazilian College of Animal Experimentation (COBEA) and Guiding Principles for Research Involving Animals and Human Beings, American Physiological Society.All efforts were made to minimize suffering.Animals were clinically monitored at least twice daily and humanely sacri ced if moribund (de ned by lethargy, dyspnea, hypothermia and weight loss).All stressed animals were sacri ced by cervical dislocation.Phenotypic assay.Plates containing solid MM were inoculated with 10 4 spores per strain and left to grow for 72 h at 37 or 44°C.When required MM was supplemented with Congo Red (10 µg/ml) or H 2 O 2 (1.5 mM).All radial growths were expressed as ratios, dividing colony radial diameter of growth in the stress condition by colony radial diameter in the control (MM at 37°C) condition.