The Aspergillus fumigatus maiA gene contributes to cell wall homeostasis and fungal virulence

In this study, two distinct in vitro infection models of Aspergillus fumigatus, using murine macrophages (RAW264.7) and human lung epithelial cells (A549), were employed to identify the genes important for fungal adaptation during infection. Transcriptomic analyses of co-incubated A. fumigatus uncovered 140 fungal genes up-regulated in common between both models that, when compared with a previously published in vivo transcriptomic study, allowed the identification of 13 genes consistently up-regulated in all three infection conditions. Among them, the maiA gene, responsible for a critical step in the L-phenylalanine degradation pathway, was identified. Disruption of maiA resulted in a mutant strain unable to complete the Phe degradation pathway, leading to an excessive production of pyomelanin when this amino acid served as the sole carbon source. Moreover, the disruption mutant exhibited noticeable cell wall abnormalities, with reduced levels of β-glucans within the cell wall but did not show lack of chitin or mannans. The maiA-1 mutant strain induced reduced inflammation in primary macrophages and displayed significantly lower virulence in a neutropenic mouse model of infection. This is the first study linking the A. fumigatus maiA gene to fungal cell wall homeostasis and virulence.


Introduction
Melanins are secondary metabolites composed of indolic monomers or phenolic polymers that act as pigments (Bayry et al., 2014).The saprophytic fungus Aspergillus fumigatus can produce two types of melanins.One of them is the dihydroxynaphthalene melanin, also known as DHN-melanin, which is one of the principal components of the conidial cell wall (Latgéand Chamilos, 2019) and together with the rodlet layer functions as the external hydrophobic layer of conidial cell wall.Underneath, forming the fibrillar layer are the b-(1,3)-glucans, chitin, a-(1,3)-glucans and galactomannan (Gastebois et al., 2009;Bayry et al., 2014;Beauvais and Latgé, 2015).When the conidia germinate, the external protective layer is shed, galactosaminogalactan is synthetized, and the fibrillar core is exposed to the environment.A second type of melanin is produced during mycelial growth, pyomelanin, which is a watersoluble pigment (Perez-Cuesta et al., 2020).This molecule is produced by spontaneous polymerization of homogentisate (HGA) during the L-phenylalanine/L-tyrosine degradation pathway (hereinafter named Phe degradation pathway).In this pathway, HGA is accumulated and converted to benzoquinone acetate by spontaneous oxidative transformation, then polymerized to pyomelanin.Finally, it is important to highlight that unlike other fungal species, such as Cryptococcus neoformans (Strycker et al., 2021) or even other Aspergillus species (Pal et al., 2013), A. fumigatus does not produce DOPA-melanin.
It has also been described that the A. fumigatus Phe degradation pathway is mediated by six genes located on chromosome 2 (hppD, hmgX, hmgA, fahA, maiA and hmgR).In addition, there are two other genes not included inside the abovementioned cluster, phhA and tat, which code for phenylalanine hydroxylase and the tyrosine aminotransferase, respectively (Heinekamp et al., 2013).They oversee the first degradation steps of Phe to Tyr, and Tyr to 4hydroxyphenylpyruvate.There are several published studies uncovering the function of the genes of the Phe/Tyr degradation gene cluster and their importance for the fungus based on deletion/ disruption mutants (Schmaler-Ripcke et al., 2009;Keller et al., 2011).However, there are still two genes of the cluster that are not well-characterized because they have not been silenced yet (fahA and maiA).
The A. fumigatus maiA is an exonic 696 bp gene (Afu2g04240) that does not contain introns in its sequence and codifies a maleylacetoacetate isomerase putatively involved in this Phe degradation pathway.Specifically, this enzyme orchestrates the isomerization of 4-maleylacetoacetate (4-MA) into 4fumarylacetoacetate (4-FA) (Perez-Cuesta et al., 2020).In theory, the resulting phenotype of the defective maiA strain should phenocopy DhmgA, which is the gene that codifies the enzyme located just upstream of maiA in this degradation pathway.
In this study, the A. fumigatus genes most strongly overexpressed upon co-culture with two different cell lines (murine macrophages and human lung epithelial cells) were studied using the AWAFUGE microarray.Among them, maiA was selected because it was overexpressed in these two in vitro models, as well as in a previously published in vivo infection (Guruceaga et al., 2018).Therefore, to study its role in fungal biology and virulence, a disruption mutant strain (maiA-1) was generated and characterized.The data obtained strongly support the involvement of maiA in the maintenance of cell wall structure as well as in the virulence of the fungus using a neutropenic mouse model.

Aspergillus fumigatus strains, media, and growth conditions
In this study, we used the A. fumigatus Af293 strain as the wild type genetic background to develop the disruption mutant strain maiA-1 as well as its complement strain maiA-1 comp.and the deletion mutant strain DmaiA.All fungal strains were cultured in Glucose Minimal Media (GMM) (6 g NaNO 3 , 0.52 g KCl, 0.52 g MgSO 4 x 7H 2 O, 1.52 g KH 2 PO 4 , 1 ml trace elements [2.2 g ZnSO 4 x 7H 2 O, 1.1 g H 3 BO 3 , 0.5 g MnCl 2 x 4H 2 O, 0.5 g FeSO 4 x 7H 2 O, 0.16 g CoCl 2 x 5H 2 O, 0.16 g CuSO 4 x 5H 2 O, (NH 4 ) 6 Mo 7 O 24 x 4H 2 O, 5 g Na 4 EDTA in 100 ml distilled H 2 O], 10 g glucose, 15 g agar, pH 6.5, in 1 liter distilled H 2 O) at 37°C for five days, and their conidia were harvested directly from the agar plates using sterile water, followed by two washes with saline solution (0.9% NaCl).The concentration of conidia needed for each experiment was calculated using a Bürker counting chamber.
To study the ability of the fungal strains to grow with a sole carbon source, GMM was replaced by a GMM medium without glucose (named salt agar).

Cell lines
The murine macrophage cell line RAW 264.7 and the human lung epithelium cell line A549, both obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA), have been used in this study.The culture conditions as well as viability calculation and passage method were done following a previously described method (Guruceaga et al., 2021).Briefly, cultures were maintained in complete RPMI (RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum, 200 mM L-glutamine, 100 U/ ml penicillin, and 0.1 mg/ml streptomycin) and incubated under standard conditions (37 °C, 5% CO 2 , and 95% humidity).Cell lines were used only when the viability of them after passage was higher than 90%.All supplies were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Phagocytosis assays and fungal behavior against RAW 264.7 cell line
To understand the behavior of the fungal strains in contact with the murine macrophages cell line RAW 264.7 we followed the method previously described (Guruceaga et al., 2021).Briefly, 2 x 10 5 cells/ml in 500 µL of cell culture RPMI (RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 200 mM Lglutamine, 100 U/ml penicillin and 0.1 mg/ml streptomycin) were seeded in 24-well plates which contained 12 mm-diameter coverslips.After overnight incubation, RAW264.7 cells were cocultured with A. fumigatus conidia at a multiplicity of infection (MOI) of 10 (ten conidia per cell) using 500 µL of challenge RPMI (RPMI 1640 medium supplemented with 10% heat-inactivated FBS, and 200 mM L-glutamine).In parallel, we seeded the same amount of conidia in the absence of RAW 264.7 cells.After 2-, 4-, 6-, and 8hours post-incubation we moved the coverslips to a new plate with cold PBS to calculate the percentage of phagocytosis, fungal germination, and hyphal branching using a Nikon Eclipse TE200-U inverted microscope.

Endocytosis assays and fungal behavior against A549 cell line
To study fungal behavior in contact with the human alveolar epithelial cell line A549 we seeded 1 x 10 6 cells/ml in 500 µL of cell culture RPMI using 24-well plates which contained 12 mm-diameter cover slips.After an overnight incubation, the cells were co-cultured with A. fumigatus conidia pre-stained with FITC (conidia were stained overnight at 4°C in an orbital shaker) at a MOI of 5 in 500 µL of challenge RPMI.At each incubation time (2, 4, 6, 8 and 10 hours), we moved the coverslip to a new plate to calculate the percentage of endocytosis.For that, a minimum of 500 cells/conidia/hyphae were counted in each replicate of the experiment using the inverted microscope Nikon Eclipse TE2000-U to calculate endocytosis (%), fungal germination (%) and hyphal branching (%) both in contact with epithelial cell lines and A. fumigatus growing alone as the control condition.

RNA isolation and purification
For RNA isolation, we collected the cells and fungus using a cell scraper after the incubation time selected in each case (6.5 hours with RAW 264.7 and 8.5 hours with A549).Then, the samples were centrifuged for 1 minute at 14,000 rpm and the pellet was suspended in 1 ml of pre-chilled DEPC sterile deionized water to lyse the mouse/ human cells.After that, the samples were centrifuged again in the same conditions abovementioned, the fungal pellet was suspended in 500 µL of pre-chilled DEPC-treated sterile deionized water and transferred to a 1.5 ml tube containing 200 µL of 0.5 mm glass beads (Sigma-Aldrich, St. Louis, MO).The samples were homogenized using the MillMix 20 beat-beater (Technica, Dulles, VA, USA) at 30 Hz for 2 minutes.Finally, we centrifuged the samples as mentioned above, and the supernatant was recovered and transferred to the extraction columns of the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany).The RNA isolation procedure was finished following the manufacturer´s instructions and the RNA quantity and integrity was verified on a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).For microarray analysis and RT-qPCR confirmation, three independent RNA samples for each time point, each of them obtained from an independent co-incubation with cells assays or controls, were studied.

Microarray selection, hybridization, and expression data analysis
The fungal transcriptome was studied using the AWAFUGE microarray (Agilent Whole A. fumigatus Genome Expression 44K v.1).The microarray hybridization and the analysis of the raw data obtained was processed as previously described (Guruceaga et al., 2018).Raw intensity data was analyzed using the limma library implemented in the Bioconductor package (Smyth, 2004;Loṕez-Romero, 2011).Background subtraction and data normalizing was done using normexp and quantile routines respectively.We fixed a statistical significance level of 0.05 after analyzing the data using ANOVA test with Benjamini-Hochberg correction.Finally, we determined the genes down-or up-regulated relative to the fungal growth alone.

Microarray data confirmation by reverse transcription quantitative PCR
We selected a subset of 22 genes to verify fungal expression profiles by RT-qPCR co-cultured with RAW 264.7 macrophages, and 22 genes with A549 pneumocytes.In both cases, we used the same 4 reference genes.Specific A. fumigatus primers were designed using Primers Quest Tool (available at " eu.idtdna.com/site") to avoid biased results due to mouse or human RNA remaining in the samples (Table S1).Selection of the most suitable housekeeping genes and analysis of the RT-qPCR results were completed following a previous publication (Guruceaga et al., 2018).
Finally, the expression levels of uge3 and pksP, two essential genes in galactosaminogalactan and DHN-melanin biosynthetic processes, were studied by RT-qPCR assays following protocols previously published (Martin-Vicente et al., 2018).

Gene target selection criteria
The selection of maiA as an important A. fumigatus gene involved in fungal virulence was determined after analyzing our transcriptomic results and previous transcriptomic studies (Guruceaga et al., 2018).We classified the genes following their fold change values (FC).We only focused our attention on those genes highly down-or up-regulated (FC < 1.5 or FC > 1.5 respectively).After that, we compared the common A. fumigatus up-regulated DEGs in three different assays: 1) A. fumigatus strain Af293 in co-incubation with lung epithelial cell line A549, 2) A. fumigatus strain Af293 in co-incubation with macrophages RAW 264.7 and 3) A. fumigatus strain Af293 infecting immunosuppressed mice (Guruceaga et al., 2018).

Mutant strains generation
The A. fumigatus mutant strains used in this study were performed following the CRISPR/Cas9 strategy previously described (Al Abdallah et al., 2017).In this study, we used the Af293 genetic background to use the same strain used in the transcriptomic studies in which we selected maiA as interesting avoiding bias interpretations.Briefly, A. fumigatus maiA mutant strains were performed using Hygromycin B (Thermo Fisher, Waltham, MA, USA) as a selection marker.For the disruption mutant strain (maiA-1), Cas9 targets and the corresponding crRNA were designed to delete the initial methionine of the maiA gene.In contrast, for DmaiA we designed two gRNAs located in the 5´UTR and 3´UTR respectively with the aim of replacing the whole target gene with the Hyg R cassette.Transformation of A. fumigatus protoplasts was carried out following the classic protocol described in the literature (Yelton et al., 1984).Finally, the complement of the maiA-1 strain (maiA-1 comp. ) was constructed by re-introducing the original maiA gene followed by the phleomycin resistance cassette in the native locus.All the primers and gRNAs used can be found in Table S2.

Determination of the pyomelanin production
Pyomelanin production was measured following a previously published protocol (Schmaler-Ripcke et al., 2009).Briefly, GMM broth (150ml) was inoculated with 1 x 10 7 conidia of each indicated A. fumigatus strain.After 20 h of pre-incubation, Phe or Tyr was added to a final concentration of 10 mM.Aliquots of 500 µL of GMM broth were taken at 24, 48 and 72 hours.Pigment formation was analyzed by direct absorbance measurements at 405 nm of the alkalized supernatants obtained by adding 20 µL of 5 M NaOH per ml of sample and centrifuging them at 16,000 x g for 2 min.Three independent replicates of the experiment were carried out on different days.

A. fumigatus cell wall stress assay, spot dilution assay and radial growth
The cell wall stressor assays were done using 6-well plates (Invitrogen, Waltham, MA, USA).The wells were filled with GMM or SMM supplemented with increasing concentrations (20,40,60,80,120, and 160 µg/ml) of CR or CFW.In parallel, plates with GMM or SMM without stressors were used as controls.To study if chitin synthase activity was affected by the maiA gene silencing, GMM 6-well plates containing (0, 0.5, 1, 2, 4, and 8 µg/ ml) of Nikkomycin Z were performed.To inoculate the plates, 5 µL of a 2 x 10 6 conidia/ml stock of each strain was pipetted at the center of the plates.The pictures were taken after incubation of the plates at 37°C for 48 h.
The spot dilution assay of each strain was done following the method previously described (Martin-Vicente et al., 2018).Briefly, we inoculated 10 4 , 10 3 , and 10 2 conidia (5 µL/drop) in GMM plates supplemented with 1.2M NaCl or 1.2M KCl as cell membrane stressor compounds or 5 mM Caffeine as a MAPK-inducing stress agent.To study the ability of the strains to use Phe, Tyr, or Phe/Tyr as the sole carbon source, we supplemented salt agar (GMM without glucose) plates with 50 mM Phe, 50 mM Tyr, or 50 mM Phe and 50 mM Tyr.The pictures were taken after incubation of the plates at 37°C for 48 h.
Furthermore, we characterized the radial growth ability of the maiA-1 mutant strain compared to the Af293.For that, we seeded a 20 µL drop containing a suspension of 10 8 /ml fresh conidia in the middle of a GMM agar plate.The plates were incubated at 37°C for five days and the radial growth of the macroscopic colonies were measured daily.
Three independent replicates of the experiment were carried out on different days.Representative pictures of each condition are shown.

Scanning electron microscopy
For the cell wall surface study, 5 x 10 5 conidia/ml from the Af293 and maiA-1 strains were seeded in 1 ml of GMM in 24-well plates containing glass coverslips.The plates were incubated at 37°C, 5% CO 2 and 95% of humidity for 12 hours.The fungal surface of each strain was studied after 2, 4, 8 and 12 hours of incubation by SEM.At each timepoint indicated, the GMM was removed, and fixing solution (2% glutaraldehyde in PBS buffer) was added to each well for 1h at 4°C . Then the samples were dehydrated through increasing ethanol concentrations and hexamethyldisilane.Finally, they were covered with gold under an argon atmosphere, and visualized under the scanning electron microscope (Hitachi S-4800).

Mouse bone marrow macrophages isolation and cytokine measurements
BMMs were isolated following a previous publication (Pellon et al., 2018).The in vitro challenge of the BMMs was done following the same conditions previously described for RAW 264.7 cells.TNF cytokine determination was doing using the mouse TNF-auncoated ELISA kit (Invitrogen, Waltham, MA, USA) following manufacturer's instructions after 16 hours of co-incubation between BMMs and A. fumigatus.

Fungal susceptibility to H 2 O 2 oxidative agent
The susceptibility of the fungal strains to H 2 O 2 was measured using GMM agar plates following a diffusion assay.Briefly, conidia of both strains (1 x 10 7 ) were seeded and spread evenly over the surface of a GMM agar plate (90 mm petri dishes).Once the fungus was seeded, a central well was generated in the middle of the agar plate using a sterile pipet tip.Finally, 50 µL of a 200 mM solution of H 2 O 2 was added to the central well.After drying for 5 minutes, the plates were incubated at 37°C for 48 hours.Three independent replicates of the experiment were carried out on different days.

Study of cell wall components
The b-glucan assay was performed following methods previously described (Fortwendel et al., 2009;Koenig et al., 2017;Souza et al., 2019).Briefly, 1 x 10 7 conidia of each strain were grown overnight into 25 ml of GMM broth at 37°C.After 16 hours of growing, hyphae were collected by filtration through Miracloth (Sigma-Aldrich, St. Louis, MO, USA) and washed using 0.1 M NaOH solution.Washed fungal hyphae were lyophilized for 24 hours.Five milligrams of dry hyphae were disrupted in a beadbeater three times (1 minute each) with 1 minute of ice incubation between each.Hyphal powder was resuspended to a final concentration of 20 mg/ml in 1 M NaOH and the solution was incubated at 52°C for 30 min.Fifty microliters of each sample were mixed with 185 µL of aniline blue staining solution (183 mM glycine, 229 mM NaOH, 130 mM HCl, and 618 mg/l aniline blue, pH 9.9) into a 96-well masked fluorescence plate (Invitrogen, Waltham, MA, USA).The sample-containing plates were incubated at 52°C for 30 min, followed by a cool down period of 30 min at room temperature.Fluorescence readings were performed using an excitation/emission wavelength of 405/460 nm respectively.All the experiments were performed in triplicate using three independent A. fumigatus cultures, and the results were represented as relative quantification versus the Af293 strain.
Chitin content analysis was performed following protocols previously published (Cańovas et al., 2017;Rocha et al., 2020).Briefly, 96 well plates containing 100 µl of GMM per well were inoculated with 10 3 conidia/well of each strain and incubated at 37 °C for 16, 18, 20, 22, and 24 hours.After each incubation time, 150 µl of CFW stock solution (10 µg/ml in cell culture PBS) was added to each well.Fluorescence (excitation, 360 nm; emission, 460nm) and absorbance (600 nm) were measured in parallel in a BioTek multiplate reader.The slope, calculated fluorescence against absorbance, of each strain was plotted.The experiment was repeated eight times with technical replicates.
Fluorescence microscopy study of cell wall chitin, mannans and exposed chitin of the fungal strains was completed using CFW (1 µg/ml in cell culture PBS), concanavalin Alexa Fluor 647 conjugate (Invitrogen, Waltham, MA, USA), and Wheat Germ agglutinin Alexa Fluor 633 conjugate (Invitrogen, Waltham, MA, USA), respectively.To do this, 24-well glass bottom plates were inoculated with 1 x 10 3 conidia and incubated for 16 hours at 37 °C.After the incubation time, samples were observed under Nikon Eclipse Ti2.

Murine model of invasive pulmonary aspergillosis
Groups of 20 female CD-1 mice (Charles River, Wilmington, MA, USA), weighing approximately 25 g, were immunosuppressed by intraperitoneal injections of 150 mg/kg cyclophosphamide (Sigma-Aldrich, St. Louis, MO) starting 4 days before the inoculation and every 3 days following, using 75 mg/kg of cyclophosphamide, and a single subcutaneous injection of 40 mg/ kg triamcinolone acetonide (Kenalog, Bristol-Myers Squibb) 24 hours before the infection.On day 0, mice were transiently anesthetized with isoflurane and challenged via nasal inoculation with 10 6 conidia in sterile saline solution.Survival was monitored at least twice a day and those animals showing severe signs of distress were humanely euthanized by anoxia with CO 2 .Survival curves were compared using the log-rank test in GraphPad Prism v. 8.2.1 for Windows.In parallel, histological studies were carried out to understand if loss of maiA impacts over the early infection establishment.To do this, 4 mice were infected with Af293, and 4 mice were infected with the maiA-1 strain, following the same protocol and conditions previously explained.At 3 days post-infection, mice were euthanized, and their lungs were harvested and fixed in 10% buffered formalin for 72 hours before being paraffin embedded.Subsequently, 4 µm slices were sectioned, mounted, and stained with hematoxylin-eosin (H&E) and Grocott's Methenamine Silver (GMS).The studies were performed in accordance to approved ethical protocols by the LACU committee of the University of Tennessee Health Science Center (Protocol number 22-0373).

Statistics
All the assays were done in triplicate on three independent days.All the statistical analysis were carried out using GraphPad Prism v. 8.2.1 (GraphPad Software Inc., San Diego, CA, USA) for Windows.In each assay, at least three biological replicates were done to measure each parameter in each condition, avoiding biased results.Error bars included in all the graphs represent standard deviation.ANOVA or t-test was used depending on if we did multiple comparisons or compared punctual data, respectively, after ensuring the data sets followed a normal distribution.

Availability of data
The ArrayExpress database (www.ebi.ac.uk/biostudies/ arrayexpress) contains the AWAFUGE microarray (v.1) design under accession number A-MEXP-2352.The same database also contains each raw microarray dataset obtained under accession number E-MTAB-13476.Moreover, the in vivo transcriptomic analysis previously published can be found in the same database under accession number E-MTAB-5314.

Study of co-incubation between A. fumigatus and macrophages
The results obtained from co-incubation with the murine macrophage RAW 264.7 cell line revealed 80% phagocytosis after 4 hours of co-incubation (Figure 1A), but no statistical differences were observed.Although the germination of conidia reached 30% at 6 hours in all cases (infection and control cultures), at the end of the experiment significantly higher germination was detected in the conidia incubated with macrophages, compared to the controls (Figure 1B).In contrast, we did not find any differences in the amount of hyphal growth or their ability to stablish a second axis of polarity during germination (Figure 1C).

Study of co-incubation between A. fumigatus and epithelial cells
Co-incubation of A. fumigatus with the lung epithelial cell line, A549, revealed a maximum endocytosis of approximately 5% after 6 hours of co-incubation (Figure 2A).In addition, the contact with these epithelial cells generated a slight inhibition of the germination rate.Indeed, 30% of germination was reached after 8 hours of coincubation with the epithelial cells while this germination rate was reached at 6 hours when the fungus grew alone (Figure 2B).However, an increase in germlings with multiple polarity axis was observed after 8 hours of co-incubation of A. fumigatus with epithelial cells (Figure 2C).

A. fumigatus gene expression in response to the co-incubation with RAW 264.7 macrophage or A549 epithelial cell lines
To understand the molecular changes that A. fumigatus suffers in the onset of the infection process, we consider 30% of germination a suitable endpoint for this purpose because in this time point both the germlings and the remaining swollen conidia are starting the activation and adaptation of their metabolism for infection.This happened after 6 and 8 hours of co-incubation with macrophages and epithelial cells, respectively, moment in which  three samples of mRNA from each condition obtained in independent experiments, as well as their respective controls, were hybridized with the AWAFUGE microarray.Once the data were analyzed and normalized, 2,137 and 5,325 A. fumigatus genes were found differentially expressed with respect to their controls (A.fumigatus without cells) when the fungus was incubated with macrophages and with human lung epithelial cells, respectively.To limit and prioritize the genes to be studied, only those displaying > 1.5 or < -1.5-fold change (log 2 ) (FC) in their conditions were considered.By this criteria, 235 A. fumigatus genes were down-regulated, and 280 genes were up-regulated during the coincubation with the macrophages (Figure 3A), whereas 534 were down-regulated and 878 genes were up-regulated during the contact with the epithelial cells (Figure 3B).
The transcriptomic data obtained from the AWAFUGE microarray was confirmed by RT-qPCR, showing a good correlation result between both techniques.Specifically, a correlation of 70.79% and 90.03% was obtained between microarray and RT-qPCR verification in the co-incubation of A. fumigatus with macrophages, and with epithelial cells, respectively (Figures 3C, D).In Figures 3E, F, FC values obtained for each gene and technique used in the validation process are shown.All of them, except those marked in red (dots in panels C and D and arrows in panels E and F), displayed a similar expression in both techniques.In addition, maiA gene expression values were plotted in green.

GO enrichment analysis of the most up/ down-regulated DEGs and comparison between different infection models
To identify those A. fumigatus processes, components, and functions of greatest importance that were impacted during the experimental infection procedure, we performed a Gene Ontology (GO) enrichment analysis (the complete analysis of DEGs can be found in Table S3).
Among the DEGs, 227 (RAW 264.7 vs Control) and 592 (A549 vs Control) were associated with a known biological process.The five most relevant processes in this category are those related with transport, regulation, response to stress, secondary metabolic process, and lipid metabolic process.
In addition, 252 A. fumigatus genes (RAW 264.7 vs Control) and 605 genes (A549 vs Control) were associated with a known molecular function.The categories corresponding to the most genes are oxidoreductase activity, hydrolase activity, and transporter and transferase activities.
Moreover, it is remarkable that 321 (RAW 264.7 vs Control) and 795 (A549 vs Control) genes have been previously associated with a known cellular location, including (in order of gene numbers associated) membrane, mitochondrion, nucleus, cytosol, plasma membrane and the extracellular region.
To select the most important genes related to infection, among the large amount of overexpressed A. fumigatus genes found, those up-regulated with a FC > 1.5 in co-incubation in common between macrophages and epithelial cells were selected.Using this approach, a total of 140 A. fumigatus genes were obtained (Table S4).These genes were then cross-referenced with those detected as overexpressed during a previous mouse infection model (Guruceaga et al., 2018).This analysis decreased the list to 13 significantly up-regulated summarized in Table 1.Among those, Afu2g04240 (maiA) gene was selected as a good candidate to be characterized in this study for its role in A. fumigatus for the next reasons: i) It is one of the genes related with the Phe degradation pathway still unstudied, ii) Its potential role in the infection process detected in our transcriptomic analysis.

The maiA gene is essential for Phe degradation pathway
The genetic manipulation strategy followed to generate the mutant strains is described in Materials and Methods section and summarized in Figure 4. Briefly, a disruption mutant was generated by replacing the initial methionine (iMet) with the hygromycin resistance gene.The resulting mutant strain (maiA-1) was resistant to hygromycin and has the target gene silenced due to the lack of the iMet (Figure 4A).The complete deletion strain (DmaiA) was also performed, through complete substitution of the maiA locus by the hygromycin resistance gene (Figure 4B).Finally, the complement of the maiA-1 strain (maiA-1 comp. ) was constructed by replacing the disrupted allele with the native maiA followed by the phleomycin resistant gene in the native locus (Figure 4C).All the strains were generated using CRISPR-Cas9 highly efficient genetic technology in the Af293 genetic background.The DmaiA strain was constructed to validate the results obtained in the disruption strain.As is possible to see in (Figures S1 and 2) the deletion strain phenocopied the maiA-1 strain results, confirming the usefulness of the disruption strategy.
As mentioned above, we hypothesized that loss-of-function mutations at the maiA locus would produce high concentrations Schematic of gene manipulations by CRISPR/Cas9 editing.(A) Silencing strategy of the maiA gene.One protospacer adjacent motif was designed to disrupt the iMet (ATG orange square) of the maiA gene.The disruption was repaired using the hygromycin resistance cassette with 40 base pair microhomology regions upstream and downstream the maiA gene in the Af293 genetic background to generate the maiA-1 mutant strain (maiA green dashed arrow represents non-functional maiA gene).(B) The complete deletion of the maiA gene was carried using two protospacers adjacent motif designed to the flanking regions of the maiA gene.The repair template process was developed following the same strategy described in panel (A) to generate the DmaiA mutant strain.(C) To complement the maiA-1 mutant strain, the native maiA gene was cloned into pAGRP plasmid (iMet included) upstream the phleomycin resistant cassette and generating the pAGRP_maiA comp.plasmid.Two protospacers adjacent motif designed to the flanking regions of the maiA gene were used.The repair template cassette was amplified from the pAGRP_maiA comp.plasmid using primers containing 40 base pair microhomology regions upstream and downstream the HygR-maiA disrupted regions, generating the maiA-1 comp.mutant strain.All the colonies were confirmed by PCR and sequencing.The genetic maps were created with BioRender.com.a Data obtained from a previous publication (Guruceaga et al., 2018).
of 4-MA that could polymerize spontaneously to pyomelanin (Figure 5A).To study the pyomelanin production ability, Af293 as well as the maiA-1 mutant and the maiA-1 comp.strains were grown for 72 hours in GMM broth (Figure 5B), GMM broth supplemented with 10 mM of Phe (Figure 5C) or GMM broth supplemented with 10 mM of Tyr (Figure 5D).Indeed, when Phe or Tyr were added to the media, the maiA-1 mutant strain accumulated pyomelanin indicated by an increase of the 405 nm absorbance signal (Figure 5C) that can be also visually observed (Figure 5D).A varying ability to degrade both amino acids was shown since the production of pyomelanin was different when Phe or Tyr were used, likely because Phe may be utilized by other metabolic pathways, compared to Tyr (Figure S3).It was also striking that when salt agar plates were supplemented with 50 mM of Phe, 50 mM of Tyr or 50 mM of both amino acids as the sole carbon source, the maiA-1 strain exhibited impaired growth (Figure 5E).Attempts to measure biomass accumulation in submerged culture to further quantify the reduced ability to utilize these amino acids as carbon sources revealed a complete lack of growth for the maiA-1 mutant in both Phe and Tyr (data not shown).

Role of maiA in growth and development of A. fumigatus
To understand the role and importance of the maiA gene in the basal growth of A. fumigatus, the maiA-1 disrupted strain was analyzed using germination and radial growth assays in GMM).Although the germination rate of all the strains was the same over the time studied (Figure 6A), slight differences in the radial growth ability were found between the Af293, the maiA-1 and maiA-1 comp.after 72 hours of growth.After 96 hours, no difference in the radial growth between Af293 and the complemented strain was found.In contrast the maiA-1 mutant strain colonies were slightly smaller compared to those from the Af293 background (Figure 6B).Although statistical differences exist at the various timepoints, no overt defects in colony formation were seen.
To better understand the role of maiA in maintaining fungal cell wall homeostasis, a stress response assay using two cell wall stressor compounds was performed (Figure 7).The maiA-1 mutant strain was hyper-susceptible to the cell wall stressors congo red (CR) and calcofluor white (CFW) as observed by the inability to grow at concentrations higher than 40 µg/ml of both compounds in GMM (Figure 7 top panels).The ability of this mutant strain to grow in the presence of both compounds was recovered by the addition of sorbitol to the media (SMM) (Figure 7 bottom panels), further supporting that growth inhibition in the mutant was due to cell wall instability.
The CFW results suggest that maiA-1 mutant could present some lack of chitin in its cell wall.However, as can be seen in Figure S4, the ability of the strains to grow in presence of Nikkomycin Z indicates that chitin synthase activity is not affected in the disruption mutant (Figure S4A).In addition, the abundance of chitin between fungal strains is the same (Figure S4B), a finding that was corroborated by fluorescent microscopy using CFW as a chitin indicator dye (Figure S4C).All these results reject the hypothesis of reduced chitin in the mutant cell wall.
The maiA-1 hyphae display an unstructured morphology Scanning electron microscopy (SEM) analysis of the mycelial appearance after culturing for 12 hours in GMM broth showed that the Af293 strain maintained a normal hyphal morphology with typical 45°branch angles and mycelia that were strongly attached to the surface (Figure 8A).In addition, analyzing the images at higher magnification, the Af293 hyphae were surrounded by an extracellular matrix, which was more prevalent at the hyphal apex (Figure 8C).
On the contrary, the maiA-1 strain displayed a disorganized mycelium, which seemed to be more aerial, less attached to the surface and in consequence more rounded (Figure 8B) without the surrounding matrix observed in Af293 (Figure 8D).
Furthermore, other cell wall components such as Nacetylglucosamine exposed residues (exposed chitin) and mannans were studied through fluorescence microscopy, and no differences either in amount (changes in fluorescence intensity) or in cell distribution along the hyphae were detected (Figures S5A, B).While chitin is more exposed in the hyphal tips and around the conidia where the first germ tube was formed (Figure S5A), mannan is homogenously distributed along the hyphae (Figure S5B) of all the fungal strains studied.Although the SEM findings could imply a lack of galactosaminogalactan due to the loss of maiA function, no differences in expression levels of uge3 between fungal strains were detected (Figure S6A).Furthermore, with the aim to exclude over production of DHN-melanin as a compensatory mechanism due to lack of maiA, expression levels of the pksP gene were measured in the fungal strains.As is possible to observe (Figure S6B), no expression differences of the pksP gene were detected between our fungal strains.
maiA is required for normal A. fumigatus virulence Finally, we characterized the pathogenic abilities of the maiA-1 strain with a set of in vitro and in vivo assays.First, we performed an in vitro characterization challenging mouse bone marrow-derived macrophages (BMMs) with our isogenic set of strains.No differences in phagocytosis rates were observed over 8 hours of co-culture between the BMMs and any of the three fungal strains (Figure 9A).These results could be related with the H 2 O 2 susceptibility assay in which similar sized zones of clearance were found in all strains (Figure 9B).It is important to highlight that this assay was developed in GMM, condition in which pyomelanin production ability of the fungal strains (Af293, maiA-1 and maiA-1 comp. ) is the same (Figure 5B).To test if pyomelanin production could have a protective effect against oxidative stress, the same assay was developed using GMM supplemented with 10 mM Phe and Tyr.As shown Figure S7, although a modest pyomelanin production can be observed (back of the plates) (Figure S7B), no strains were inoculated onto salt agar plates or salt agar plates in which the only carbon source was Phe (50 mM), Tyr (50 mM), or Phe and Tyr (50 mM each).Plates were incubated at 37°C for 72h.Note that maiA-1 mutant showed a non-significant basal growth probably produced by residual sugars present in the agar.No growth was detected when the assay was done in liquid media following the same conditions (Data not shown).
statistical differences regarding the size of inhibition could be detected (Figure S7A).
In a detailed study of the BMM cells-pathogen interaction, significantly less TNF release was observed in the macrophages cocultured with the maiA-1 mutant strain compared to those cocultured with either the wild-type or maiA-1 comp.strain (Figure 9C).A very interesting finding was discovered when the b-glucan content of each fungal strain was studied.Specifically, the total glucan content was found to be significantly lower in the maiA-1 mutant strain compared to the wild-type strain (Figure 9D).This result might indicate that reduced stimulation of BMMs induced by the maiA-1 mutant strain could be attributable to its lower bglucan content.
Moreover, the impact of the absence of the maiA gene on virulence was studied using a neutropenic mouse model of invasive aspergillosis (IA) (Figure 9E).The results revealed that 85% and 90% of the mice infected with the Af293 strain and the maiA-1 comp., respectively, had died after 15 days post-infection.In contrast, the Cellular characterization results of the maiA-1 mutant strain.(A) Percentage of germination during 8 hours of incubation in GMM of the Af293, maiA-1, and maiA-1 comp.strains.(B) Colony diameters of 2 x 10 6 fresh conidia of each strain (Af293, maiA-1, and maiA-1 comp. ) seeded onto GMM plates.The diameter of each strain was measured daily for a maximum of 5 days.Statistical analyses were performed by two-way ANOVA with Tukey's test for multiple comparisons.*p< 0.05.

A B
Phenotypic characterization of Af293, maiA-1, and maiA-1 comp.strains using GMM agar 6-well plates.GMM or SMM 6 well plates supplemented with different concentrations (0, 20, 40, 60, 80, 120, 160 µg/ml) of congo red (CR) (A) or calcofluor white (CFW) (B) were inoculated using 1 x 10 3 fresh conidia of each strain (Af293, maiA-1, and maiA-1 comp.).The plates were incubated at 37°C for 48 h, time in which pictures were taken.Identical control wells are demonstrated in the two upper panels for GMM alone without the addition of either CR or CFW, and identical control wells are demonstrated in the two lower panels for SMM alone.Representative images of 3 replicates per condition are shown.
group of mice infected with the maiA-1 strain showed 40% survival at the end of the assay.The overall statistical analysis of the assay showed a significantly lower lethality of the maiA-1 mutant strain.This lower lethality showed by the maiA-1 strain is not explained by the ability of this mutant to establish infection in lungs.As seen in Figure 10, no overt differences in fungal lesion presentation were found between those mice infected with the Af293 and the maiA-1 strain at 3 days post-infection.Regarding the inflammation produced by the strains (Figure 10A), the lungs from mice infected with the Af293 strain presented a slightly higher inflammation homogeneously distributed through the organ than those lungs infected with the maiA-1 strain.In relation with the GMS images (Figure 10B), is easy to see multiple A. fumigatus foci of both strains growing in the lungs after 3 days post-infection (1.5 magnification), but as more magnifications the pictures were taken, is easier to find more Af293 foci infecting the lung parenchyma than the maiA-1 fungal strains that is more located around the airways and less in the parenchyma.Either way, these slight differences found between both fungal strains during the early stages of infection establishment likely do not explain the mortality differences.

Discussion
In this work, RAW 264.7 macrophages and A549 human lung epithelial cells were used as in vitro infection models in coincubation with A. fumigatus to perform two independent transcriptomic studies.After characterization of both models during co-incubation, when A. fumigatus reached 30% germination, the total mRNA was isolated and hybridize with the AWAFUGE 1.0 microarray.The transcriptomic analysis of both in vitro experimental assays showed that 140 A. fumigatus were upregulated (FC > 1.5) in common in the two models used.To further prioritize the number of genes of interest, they were then compared with those detected as up-regulated during a previously published murine intranasal infection (Guruceaga et al., 2018).Using this approach, 13 genes were found to be up-regulated in all three abovementioned infection procedures.Some of these genes have been previously studied by other authors and have been related with the infection process or detected as expressed at the onset of the contact between cells or sera.Among them, it is important to highlight genes such as aspf1, that codifies an A. fumigatus allergen commonly found in different proteomic/transcriptomic studies (Guruceaga et al., 2018;Ramirez-Garcia et al., 2018) or the fumarate reductase Osm1, whose expression has been previously observed during hypoxia and during the first contact between the dormant conidia and the host (Teutschbein et al., 2010;Vödisch et al., 2011).Two up-regulated genes codify TCA cycle enzymes: the succinate dehydrogenase Sdh1 and the citrate synthase Cit1.The former protein has been previously described as expressed after human neutrophil exposure (Sugui et al., 2008), during the exit of the dormancy (Lamarre et al., 2008) as well as under hypoxia (Vödisch et al., 2011).Regarding the Cit1, some researchers have detected its expression during the contact with sera (Asif et al., Characterization of the Af293, maiA-1, and comp.strains during the infection process.(A) Phagocytosis measured after 8 hours of coincubation between the three fungal strains and primary BMMs.To do that, 500 µL of challenge RPMI containing 2 x 10 5 BMMs and A. fumigatus conidia in a MOI of 10 were seeded in 24-well plates which contained 12 mm-diameter coverslips.(B) Inhibition halus diameter of a culture of the three fungal strains on GMM agar plates after 48 hours of incubation in presence of 200 mM H 2 O 2 in a central well of the plate.To do that, 1 x 10 7 fresh conidia of each strain were seeded and spread over a GMM aga plate.After that, 50 µL of a 200 mM solution of H 2 O 2 was added into a well previously generated.The plates were incubated at 37°C for 48h.(C) BMMs TNF release after 16 hours challenge with the three fungal strains following the conditions described in panel 2010) and with airway epithelial cells (Oosthuizen et al., in addition, the cit1 gene was also described as essential for IA manifestation (Ibrahim-Granet et al., 2007).
The A. fumigatus catabolic Phe degradation pathway is encoded by a cluster of 6 genes (hppD, hmgX, hmgA, fahA, maiA, and hmgR) in this microorganism (Schmaler-Ripcke et al., 2009;Perez-Cuesta et al., 2020).Except for two genes (fahA and maiA), the cluster of genes involved in this pathway has been previously studied through the development of loss-of-function mutant strains.The DhppD null mutant displayed a lack of pyomelanin synthesis because HGA formation was blocked.Furthermore, the hyphae of this mutant showed increased susceptibility to hydrogen peroxide and other oxidizing agents when compared with Af293 (Schmaler-Ripcke et al., 2009).Moreover, this gene has been detected as overexpressed at 96 hours post-intranasal infection in comparison with the first 24 hours in immunosuppressed mice (Guruceaga et al., 2018).Other authors also detected the overexpression of hpdA, the hppD homolog in Penicillium marneffei, after exposure to macrophages (Boyce et al., 2015).In addition, the DhmgX mutant strain has been shown to have similarities with a DhppD strain (Keller et al., 2011).This DhmgX mutant was unable to produce HGA and, therefore, HmgX could be a cofactor or mediator of HppD enzyme function (Keller et al., 2011).On the other hand, the DhmgA mutant strain showed an increase in pyomelanin owing to the accumulation of HGA and displays reduced susceptibility to oxidizing agents compared to DhppD (Schmaler-Ripcke et al., 2009).Finally, the last gene of the Phe degradation pathway studied was the transcription factor hmgR.The DhmgR mutant is incapable of using Tyr as the sole carbon source or nitrogen source and, therefore, only a residual accumulation of pyomelanin was observed in cultures (Keller et al., 2011).
The only published study concerning maiA in A. fumigatus showed that this gene was overexpressed when the medium was supplemented with Tyr (Keller et al., 2011).Orthologs of this gene has been silenced in both Aspergillus nidulans and P. (Fernańdez-Cañón and Peñalva, Fernańdez-Cañón and Peñalva, 1998;Boyce et al., 2015).The catabolic Phe degradation pathway, in which maiA is putatively involved, is responsible for the production of one type of soluble melanin, known as pyomelanin, when an accumulation of intermediate metabolites, such as HGA or 4-MA, occurs.The mutation of the maiA gene is predicted to result in truncation of the final part of the pathway inhibiting the degradation of Phe or Tyr to fumarate and acetoacetate, which are destined to the Krebs cycle or the biosynthesis of other amino acids.Consequently, there would be an accumulation of 4-MA and HGA and the subsequent production of pyomelanin.
In accordance with previous publications (Schmaler-Ripcke et al., 2009;Keller et al., 2011;Perez-Cuesta et al., 2020), in response to Phe and Tyr, the maiA-1 mutant strain would block the Phe degradation pathway causing a probable accumulation of 4-MA and, likely, of the previous metabolites of the pathway that can polymerizate to pyomelanin.Specifically, the mutant strain maiA-1 started to produce this pigment when Phe and, especially, Tyr were added to the medium as the only carbon source.These results support our hypothesis that A. fumigatus can uses in a more effective way Phe than Tyr as an alternative carbo source.However, results obtained by other investigators using the DhmgA mutant showed that this strain produced more pyomelanin by the addition of Phe than Tyr (Schmaler-Ripcke et al., 2009).
Regarding basal growth analysis, our results suggest that the absence of the maiA gene did not impact on germination and minimally impacted on basic radial growth of the colonies.In contrast, the cell wall stress assays carried out in this study pointed out that the maiA-1 strain suffers an alteration of the cell wall stability, which makes it unable to grow in presence of CR or CW, perhaps due to the lower content of glucans detected in its wall.Although, CR and CW are compounds with a chemical structure that can interact with b-linked glucans and, specifically, it is thought that both compounds interfere with cell wall assembly by binding to chitin, no differences in chitin content were detected between the fungal strains.Conceivably, CW and CR act by binding to nascent chitin chains, thereby inhibiting the assembly enzymes that connect chitin to b-1,3-glucan and b-1,6-glucan (Herth, 1980;Wood, 1980;Pringle, 1991).The ability to grow in the presence of both compounds was recovered by osmotic stabilization using sorbitol.
Other studies performed with mutant strains related with highosmolarity glycerol response pathway (shoA, msbA, opyA, sakA and mpkC) which reported similar cell wall phenotypes (i.e., more sensitive to cell wall stress) are also more sensitive to osmotic stresses (Mattos et al., 2020;Silva et al., 2020) and to caspofungin.In contrast, the maiA-1 strain displayed the same growth ability as the Af293 strain in the presence of osmotic stresses such as KCl or NaCl.In addition, the maiA-1 strain showed normal susceptibility to the antifungal drugs tested (voriconazole, fluconazole, posaconazole, caspofungin and micafungin) following the EUCAST standardized method (data not shown).It is true that, while the maiA gene is involved in Phe degradation pathway, the other aforementioned genes are involved in the high-osmolarity glycerol response pathway and, therefore, the studies are not totally comparable.Another possible explanation is that, although the maiA-1 mutant strain displayed a cell wall defect, this is not caused through an issue in the cell wall integrity pathway because this mutant strain was not affected by the presence of caffeine in the media as other investigators have previously described (Valiante et al., 2008;Valiante et al., 2009).This issue could be generated by a blockage of the Phe degradation pathway since the maiA-1 strain is unable to complete the catabolic process of Phe and Tyr by which the fungus could synthesize new amino acids for structural purposes.
The SEM analysis showed a general unstructured appearance of the maiA-1 mutant strain confirming a cell wall issue as we hypothesized from the CR and CFW stress assays, that is not related with the amount or distribution of chitin, exposed chitin or mannans in the cell wall.The Af293 hyphae looked similar to other SEM images presented by other investigators previously (Wuren et al., 2014;Gonzaĺez-Ramıŕez et al., 2016;Joubert et al., 2017).However, the maiA-1 hyphae showed differences in the external structure of the cell wall and a lack of a putative matrixlike substance that surrounded the apex of Af293 hyphae, although no differences in transcriptional levels of uge3 were detected and in consequence a lack of galactosaminogalactan (Gravelat et al., 2013).
The last step of the research was to study the role of the maiA gene in fungal virulence.The cell wall is the main fungal target recognized by the immune cells and, for that, an in vitro model of infection using primary BMMs was performed.The maiA-1 and Af293 strains showed the same H 2 O 2 resistance ability, similar to the results previously described using DhppD and DhmgA mutants (Schmaler-Ripcke et al., 2009).Besides, Boyce and coworkers performed the same experiment with P. marneffei also found that DhppA and DhmgA showed a very mild sensitivity to H 2 O 2 (Boyce et al., 2015).No other studies have previously described a similar phenotype with alterations in cell wall, and b-glucans composition, upon silencing genes involved in the Phe degradation pathway in A. fumigatus, A. nidulans or P. marneffei.However, this result could explain, not only why the maiA-1 strain produced significantly less TNF release by BMMs after 16 hours, but also the phenotype showed in response to CR and CFW and the SEM images.
The lower mortality observed in neutropenic mice infected with the mutant strain is novel and interesting because other mutant strain of a gene involve in the Phe degradation pathway (DhppD) did not show any difference in the mortality rate observed when was compared with its complemented strain (hppDc) (Keller et al., 2011).In addition, it is important to note that these other studies used only transient immune suppression with corticosteroids instead of our model (cyclophosphamide and triamcinolone acetonide) that employs persistently immunosuppressed mice.What we can conclude about the lower mortality of the maiA-1 strain is that is not caused by an issue in the establishment of the infection; more studies focused on how loss of maiA can modulate or impacts host-pathogen response are needed to understand how this gene interferes with the mortality ability.
In this study, an involvement of maiA with fungal cell wall, which is seemingly independent of classical MAPK regulation, has been established.Furthermore, a new Phe/Tyr fitness theory by which the use of Phe as sole carbon source is energetically more beneficial to A. fumigatus is proposed.Finally, the role of the maiA gene in fungal virulence has been revealed using a chemotherapeutically immune suppressed murine infection model.In fact, this the first study which a mutation of one of the genes involved in the Phe degradation pathway is directly related to A. fumigatus virulence.The process by which the maiA-1 strain suffers cell wall defects should be studied in future investigations, but the impossibility of this mutant strain to utilize these amino acids in a more efficient way could be involved in the phenotype observed both related with cell wall defects as well as the decreased virulence.
plotting CFW fluorescence against biomass measured as absorbance at nm.The results the average ± SEM (n=16).C) Fluorescent micrographs taken after 16 hours culture at 37°C and stained with a CFW stock solution (1 µg/ml in PBS) for 10 minutes in dark.Samples were taken using a Nikon Eclipse Ti2 microscope.To make easy the comparison of fluorescence between samples all the pictures were taken the same day and using 2 milli seconds of exposition time.In panels A and C, representative images of 3 replicates per condition are shown.

SUPPLEMENTARY FIGURE 5
Loss of maiA function does not impact the amount and distribution of exposed chitin or mannans in the cell wall.Fluorescent micrographs taken after 16 hours culture at 37°C and stained with a mix of CFW and WGA-Alexa Fluor 633 conjugate (A), or ConA-Alexa Fluor 647 conjugate (B).In panel A, we first show chitin staining with CFW to watch the whole hyphae, then exposed chitin stained with WGA and finally the co-localization of both after merging both images.To make easy the comparison of fluorescence between samples panel A pictures were taken the same day and using 2 milli seconds of exposition time for the CFW and 20 milli seconds for WGA, and panel B pictures were taken using 9 milli seconds of exposition time.Representative images of 3 replicates per condition are shown.
FIGURE 1 Characterization of the co-incubation between the fungal strain Af293 and the macrophage cell line RAW 264.7.(A) Phagocytosis assay during 8 hours of co-incubation with Af293.To do that, 500 µL of challenge RPMI containing 2 x 10 5 macrophages and A. fumigatus conidia in a MOI of 10 were seeded in 24-well plates which contained 12 mm-diameter coverslips.(B) Percentage of germination of the Af293 strain alone and in with coincubation with RAW 264.7 macrophages.The dashed line indicates 30% germination.(C) Percentage of Af293 germlings with multiple polarity axes alone and in co-incubation with RAW 264.7 macrophages.Statistical analysis in panel (A) was performed by One-Way ANOVA followed by multiple comparisons.ns p > 0.05.Statistical analysis in panel (B, C) were performed by t-test.*p < 0.05.
FIGURE 2 Characterization of the co-incubation between the fungal strain Af293 and the human epithelial cell line A549.(A) Endocytosis assay during 10 hours of co-incubation with Af293.To do that, 500 µL of challenge RPMI containing 1 x 10 6 macrophages and A. fumigatus conidia in a MOI of 5 were seeded in 24-well plates which contained 12 mm-diameter coverslips.(B) Percentage of germination of Af293 in co-incubation with the lung epithelial cell line A549 in the same conditions explain in panel (A).The dashed line indicates 30% germination.(C) Percentage of Af293 germlings with multiple polarity axes alone and in co-incubation with the lung epithelial cell line A549 in the same conditions explain in panel (A).Statistical analysis in panel (A) was performed by One-Way ANOVA followed by multiple comparisons.ns p > 0.05.Statistical analysis in panel (B, C) were performed by t-test.*p < 0.05.
FIGURE 3 Gene expression analysis.Volcano plot showing A. fumigatus differentially (black) (FC > 1.5 or FC < -1.5) and non-differentially (grey) expressed genes in co-incubation with RAW 264.7 macrophages (A), A549 lung epithelial cells.(B).The x-axis values represent the fold change (log 2 ) of microarray data and the y-axis values represent the statistical significance (-log 10 P).Spots with positive values indicate upregulation of the gene during coincubation with the indicated cell line.(C, D) Correlation analysis between microarray and RT-qPCR data.The x-axis values represent the fold change (log 2 ) of microarray data and the y-axis values represent the fold change (log 2 ) of RT-qPCR results of the selected fungal genes in coincubation with RAW 264.7 macrophages (C), and A549 lung epithelial cells (D).Each point corresponds to the mean value from three independent samples.(E) Comparative levels of fungal gene expression between the microarray and the RT-qPCR (Af293 in co-incubation RAW 264.7 macrophages).(F) Comparative levels of fungal genes expression between the microarray and the RT-qPCR (Af293 in co-incubation with A549 lung epithelial cells).Green points in (A-D) and green arrows in (E, F) show the corresponding results obtained for maiA.Red spots in (C, D) and red arrows in (E, F) show contradictory results in gene expression detected between microarrays and RT-qPCR results.
FIGURE 5 Fungal strain characterization results in relation to pyomelanin production and growth ability in presence of Phe and/or Tyr.(A) Phe degradation pathway including the genetic cluster and the proteins involved in each step.The red cross indicates the point in which the pathway is disrupted due to the disruption produced in maiA-1 strain.The diffused grey area indicates the part of the pathway that the maiA-1 strain cannot perform.Pyomelanin secretion time course of Af293, maiA-1, and maiA-1 comp.strains growing for 72 hours in (B) GMM broth, (C) GMM broth supplemented with 10 mM Phe and (D) GMM broth supplemented with 10 mM Tyr. Dashed boxes in (B-D) indicate quantitative information corresponding to the photographs indicated by arrows.Statistical analyses were performed by Two-way ANOVA with Tukey's test for multiple comparisons.*p < 0.05 or **p < 0.001.(E)Ten thousand fresh conidia of the Af293, maiA-1, and maiA-1 comp.strains were inoculated onto salt agar plates or salt agar plates in which the only carbon source was Phe (50 mM), Tyr (50 mM), or Phe and Tyr (50 mM each).Plates were incubated at 37°C for 72h.Note that maiA-1 mutant showed a non-significant basal growth probably produced by residual sugars present in the agar.No growth was detected when the assay was done in liquid media following the same conditions (Data not shown).

FIGURE 8
FIGURE 8 Scanning electron micrographs of A. fumigatus strains growing in GMM broth for 12 hours.(A, C) SEM images of Af293 hyphae.(B, D) SEM images of maiA-1 hyphae.Red arrows denote specific details described in text.Representative images of 3 replicates per condition are shown.
(A). (D) Relative content of b-glucans quantified in the three fungal strains.(E) Kaplan-Meier analysis of chemotherapeutically immune suppressed infected mice for 15 days post intranasal exposition to Af293, maiA-1, and maiA-1 comp.fungal strains.Statistical analyses of panels (C, D) were performed by Two-way ANOVA with Tukey's test for multiple comparisons.*p< 0.05.Statistical analysis of panel (E) was done by Log-Rank test.*p< 0.05; **p< 0.001.
FIGURE 10Histological study of mice infected with Af293 and maiA-1 strain after 3 days post-infection.(A) Lung sections stained with Hematoxylin -Eosin (H&E).Pictures were taken with 5x magnification after 3 days post-infection.Black arrows indicate inflammation foci through the lung tissue.(B) Lung sections stained with Grocott's Methenamine Silver (GMS).In the figure is possible to see consecutive magnifications from 1.5x to 23x of the same section of the lungs.Red squares highlight the area amplify in the next picture.The most representative pictures were shown in the figure.

TABLE 1
Common overexpressed DEGs in the three experimental infection models.