Functional Characterization of an Aspergillus fumigatus Calcium Transporter (PmcA) that Is Essential for Fungal Infection

Aspergillus fumigatus is a primary and opportunistic pathogen, as well as a major allergen, of mammals. The Ca+2-calcineurin pathway affects virulence, morphogenesis and antifungal drug action in A. fumigatus. Here, we investigated three components of the A. fumigatus Ca+2-calcineurin pathway, pmcA,-B, and -C, which encode calcium transporters. We demonstrated that CrzA can directly control the mRNA accumulation of the pmcA-C genes by binding to their promoter regions. CrzA-binding experiments suggested that the 5′-CACAGCCAC-3′ and 5′-CCCTGCCCC-3′ sequences upstream of pmcA and pmcC genes, respectively, are possible calcineurin-dependent response elements (CDREs)-like consensus motifs. Null mutants were constructed for pmcA and -B and a conditional mutant for pmcC demonstrating pmcC is an essential gene. The ΔpmcA and ΔpmcB mutants were more sensitive to calcium and resistant to manganese and cyclosporin was able to modulate the sensitivity or resistance of these mutants to these salts, supporting the interaction between calcineurin and the function of these transporters. The pmcA-C genes have decreased mRNA abundance into the alveoli in the ΔcalA and ΔcrzA mutant strains. However, only the A. fumigatus ΔpmcA was avirulent in the murine model of invasive pulmonary aspergillosis.


Introduction
Calcium ions are extremely important for signal transduction. Two important calcium mediators in the eukaryotic cell are calmodulin and the phosphatase calcineurin [1,2]. Calcineurin is a heterodimeric protein composed by a catalytic subunit A and a regulatory subunit B [1]. In fungi, calcineurin plays an important role in the control of cell morphology and virulence [1,2,3,4]. The main mode of action of calcineurin is through the dephosphorylation of the transcription factor Crz1p [5]. Calcineurin dephosphorylates Crz1p upon an increase in cytosolic calcium, allowing its nuclear translocation [5,6]. CRZ1 deficient mutants display hypersensitivity to chloride and chitosan, a defective transcriptional response to alkaline stress and defects in cellular morphology and mating [5,7,8,9]. Inactivated Schizosaccharomyces pombe CRZ1 mutants (Dprz1) are hypersensitive to calcium and have decreased transcription of the Pmc1 Ca +2 pump [10]. C. albicans homozygotes crz1D/D display moderately attenuated virulence and sensitive to calcium, lithium, manganese, and sodium dodecyl sulfate [6,11,12].
We and others have been characterizing the Ca +2 -calcineurin pathway in the human pathogenic fungus A. fumigatus [3]. In this fungus calcineurin is need for hyphal extension, branching and conidial architecture. Furthermore, the A. fumigatus DcalA mutant strain has decreased fitness in a low dose murine infection, cannot grow in fetal bovine serum (FBS), and is deficient in inorganic phosphate transport [3,13]. Three other elements in this pathway were also characterized: (i) the transcription factor CrzA [14,15], (ii) the RcnA/CbpA, belonging to a class of endogenous calcineurin regulators, calcipressins [16,17], and (iii) the Golgi apparatus Ca +2 /Mn +2 P-type ATPase PmrA [16]. CrzA mediates cellular tolerance to increased concentrations of calcium and manganese [14,15]. In addition to acute sensitivity to these ions and decreased conidiation, the crzA null mutant suffers from decreased expression of calcium transporters under high calcium concentrations and a loss of virulence. The last identified component of the pathway in A. fumigatus, PmrA, has been demonstrated to play a role in cation homeostasis and in the cell wall integrity pathway [16].
Fungal vacuolar Ca 2+ ATPases are involved in removing Ca 2+ ions from the cytosol and transporting them to internal stores thus avoiding calcium toxicity [18]). In fungi, the vacuole is a major calcium store and the two main pathways that facilitate the accumulation of Ca +2 into vacuoles are the Ca +2 -ATPases and Ca +2 /H + exchangers [18]. In S. cerevisiae, PMC1 is responsible for this process preventing growth inhibition by the activation of calcineurin in the presence of elevated calcium concentrations [19]. Here, we report the molecular characterization of three A. fumigatus PMC1 calcium transporter-encoding genes, pmcA-C. We demonstrated that CrzA directly controls the pmcA-C mRNA accumulation via binding to their promoter regions. We constructed null mutants for pmcA-B, a conditional mutant for pmcC and investigated the phenotypes/virulence of these deletions in a murine model of invasive pulmonary aspergillosis. We show that A. fumigatus pmcC is an essential gene, while pmcA and pmcB are both involved in calcium and manganese metabolism. However, only pmcA had a dramatic impact on A. fumigatus virulence and pathogenicity, since A. fumigatus DpmcA was avirulent in a murine model of invasive pulmonary aspergillosis.

Identification of three A. fumigatus PMC1 homologues
The three main calcium transporters responsible for calcium metabolism in S. cerevisiae are PMC1, VCX1, and PMR1 [20]. A phylogenetic analysis was performed in order to learn more about homologues of these transporters and other putative A. fumigatus calcium transporters (Figure 1). Previously we observed that the mRNA abundance of two PMC1 orthologous genes, pmcA (Afu1g10880) and pmcB (Afu3g10690), which encode calcium transporters, was dependent on CalA and CrzA . By using this approach, we have identified an additional PMC1 orthologue, pmcC (Afu7g01030). S. cerevisiae VCX1 encodes a vacuolar antiporter with Ca +2 /H + and K + /H + exchange activity, which is involved in the control of cytosolic Ca +2 and K + concentrations [21]. There are four A. fumigatus Vcx1p homologues, Afu1g04270 and Afu4g03320 (possibly paralogues), Afu2g07630 and Afu2g05320 ( Figure 1). Finally, S. cerevisiae PMR1 encodes a high affinity Ca +2 /Mn +2 P-type ATPase required for Ca +2 and Mn +2 transport into the Golgi [22]. We have identified two A. fumigatus PMR1 homologues, Afu2g05860 and Afu6g06740 ( Figure 1). Recently, A. fumigatus pmrA (Afu2g05860) was characterized [16]. The DpmrA mutant strain has increased bglucan and chitin content and it is hypersensitive to cell wall inhibitors, but remains virulent. In addition to these three classes of transporters, we also identified homologues for the calcium channel subunit Mid1 (Afu5g05840), an H + /Ca +2 exchanger (Afu2g05330), the calcium channel subunit Cch1 (Afu1g11110), and a calcium permease family membrane transporter ( Figure 1).
Here, we concentrate our attention on the molecular characterization of A. fumigatus PMC1 homologues. These three putative proteins showed approximately 45% identity and 67% similarity (e-value from 7.0e-160 to 1.4e-208) to the S. cerevisiae PMC1 homologue. PmcA demonstrates 51% identity and 63% similarity with PmcB (e-value 6.9e-271) and 45% identity and 58% similarity with PmcC (e-value 4.e-174) while, PmcB and PmcC showed 53% identity and 66% similarity (e-value 2.7e-276) (for the Clustal aligment of these three proteins, see Supplementary Figure S1). PmcA-C are closely related and probably paralogues ( Figure 1). In addition to pmcA and pmcB, the pmcC gene also has decreased mRNA abundance in the DcalA and DcrzA mutant strains, respectively, when exposed in vitro to CaCl 2 200 mM, compared to wild-type A. fumigatus [14,17, data not shown]. To address if CrzA is directly controlling the transcription of pmcA-C, we performed Electrophoretic Mobility Shift Assays (EMSA) using purified recombinant GST::CrzA produced in E. coli. Previously, we performed an in silico analysis using MEME (Motif-based sequence analysis tools; http://meme.sdsc.edu/meme4_1_1/ intro.html) to detect the possible presence of a calcineurindependent response elements (CDREs)-like consensus motifs in the promoter regions of 28 A. fumigatus CrzA-dependent genes [17]. By analyzing their promoter regions, 59-GT[T/G]G[G/C][T/ A]GA[G/T]-39 was defined as the CDRE-consensus sequence for A. fumigatus AfCrzA-dependent genes. When the pmcA-C promoter regions (about 500-bp upstream ATG) were scanned for putative CDRE motifs, we were able to identify the 59-CCCTGCCCC-39 and 59-CACAGCCAC-39 sequences (at 2156 and 2102 bp from the ATG start codon, respectively) in the pmcA and pmcC promoter regions. However, we could not identify any conserved CDRE motif in the pmcB promoter region (Supplementary Figure S2). Three DNA fragments of about 300-bp located upstream the putative ATG initiation codon of pmcA-C genes were used as probes (Supplementary Figure S2). DNAprotein complexes with reduced mobility were observed in the three DNA fragments (Figure 2), however the complexes affinities were different among the three fragments. While 2 mg of GST::CrzA were required for the binding of the pmcC probe ( . This result suggests a strong CrzA affinity for this DNA fragment. The specificity of the DNA-protein complex was also confirmed by using mutated pmcA and pmcC probes, in which the core sequences were changed by site-directed mutagenesis. We have not investigated a mutated pmcB DNA fragment because we were not able to identify a conserved CDRE motif in this upstream region. We have not observed the formation of any complex by using both mutated DNA fragments as probes ( Figure 2, lane 9 for mpmcA probe and lane 23 for mpmcC probe). An interesting result was the presence of two complexes exhibiting different molecular masses for pmcA and pmcB probes. We speculate that they may represent complexes with distinct conformational structures. Additional experiments will be necessary to clarify this. Taken together our results suggest that the mRNA accumulation of pmcA-C is directly regulated by CrzA.

Construction of the A. fumigatus pmcA-C mutants
To get a greater understanding of the role of pmcA-C, we tried to inactivate all three genes (Supplementary Figure S3). However, we were unable to inactivate pmcC, suggesting that this is an essential A. fumigatus gene. Thus, we constructed an alcA::pmcC mutant by replacing the endogenous pmcC promoter with the alcA promoter and verified its growth when the alcA promoter was repressed. The alcA promoter is repressed by glucose, derepressed by glycerol and induced to high levels by ethanol or L-threonine [23]. We selected a transformant that when transferred from 16 h growth in 2% glycerol, as single carbon source, to 2% glycerol +100 mM threonine for 6 h, the mRNA accumulation of pmcC was approximately 15-fold higher than when grown in the presence of 4% glucose ( Figure 3A). The repression of alcA by growing the alcA::pmcC mutant strain in the presence of 4% glucose decreased colony diameter dramatically ( Figure 3B). In contrast, both wildtype and alcA::pmcC strains demonstrated similar radial diameter when grown in 2% glycerol ( Figure 3B). Interestingly, pmcC overexpression also decreased the colony diameter size when compared to the wild-type strain, suggesting increased PmcC activity causes some metabolic disturbance that affects growth ( Figure 3B). These results strongly indicate pmcC is an essential A. fumigatus gene.
We also compared the absolute levels of mRNA abundance among pmcA, -B, and -C when the A. fumigatus wild-type, DpmcA and DpmcB mutant strains were exposed to 200 mM CaCl 2 ( Figure 4). Upon exposure of wild-type A. fumigatus to calcium, pmcB mRNA levels were higher than pmcA and pmcC, while pmcA levels were higher than pmcC ( Figures 4A-C). The number of normalized pmcA and pmcC transcripts in the DpmcB and DpmcA mutant strains, respectively, were not different from the wild-type strain ( Figures 4A and B), suggesting the absence of either pmcB or pmcA does not considerably affect the mRNA abundance of pmcA and pmcC. However, before adding 200 mM CaCl 2 there was approximately six times more pmcB transcripts in the DpmcA than in the wild-type strain ( Figure 4B). Interestingly, the pmcC mRNA levels are reduced upon CaCl 2 exposure in both DpmcB and DpmcA mutant strains. These results suggest that there is compensation in the mRNA accumulation of pmcB in the DpmcA mutant strain and pmcC mRNA accumulation is dependent on pmcA and pmcB. Upon calcium exposure, down-regulation or overexpression of pmcC had no effect on pmcA or pmcB mRNA accumulation (data nor shown).
Next, we characterized the phenotype of DpmcA and DpmcB by growing these strains in different culture media in the presence and absence of cyclosporin A (CsA). This immunosuppressive drug inhibits calcineurin signaling by forming a complex with the immunophilin cyclophilin which then inhibits calcineurin [24]. In addition, since the DcrzA mutant is also sensitive to MnCl 2 , we decided to investigate a possible influence of pmcA-B on this phenotype. Curiously, the DpmcA mutant strain demonstrated different behavior in complete (YAG) and minimal media (MM) ( Figure 5A). It showed reduced radial growth rate in complete medium when compared to the wild-type strain, but this reduction in growth was not suppressed by cyclosporin 25 ng/ml (Figures 5A). The DpmcA mutant strain was sensitive to CaCl 2 500 mM and showed increased sensitivity in YAG and MM, compared to both the wild-type and other mutant strains, when cyclosporin 25 ng/ml was added ( Figure 5B). The DpmcA mutant strain was resistant to MnCl 2 25 mM in both YAG and MM media, however cyclosporin suppressed DpmcA resistance in YAG and wild-type sensitivity in MM ( Figure 5C). The DpmcB mutant strain had about the same radial diameter than the wild-type strain in both MM and YAG media ( Figure 6A), but it was much more sensitive to CaCl 2 in YAG and showed increased sensitivity when grown in the presence of cyclosporin ( Figure 6B). However, in MM+500 mM CaCl 2 the DpmcB mutant strain has the same radial diameter as the wild-type strain ( Figure 6B). In addition, the DpmcB mutant strain was more resistant to YAG+25 mM MnCl 2 than the wild-type strain ( Figure 6C), but this resistance was suppressed in the presence of cyclosporin ( Figure 6C). The same growth was observed for both wild-type and DpmcB when grown in A. fumigatus has three S. cerevisiae PMC1 homologues. Phylogram tree and multiple sequence alignment of calcium transporter orthologues were made in CLUSTAL W2 (http//www.ebi.ac.uk/Tools/clustalw2/index.html) using the default parameters. The followings proteins were used for the analysis: Afu3g10690 (pmcB; XP_754550); Afu7g01030 (pmcC; XP_746828); Afu1g10880 (pmcA; XP_752453); Afu2g05860 (calcium/ mangenese P-type ATPase: XP_749715); Afu6g06740 (endoplasmic reticulum calcium ATPase: XP_750567); Afu5g05840 (calcium channel subunit Mid1: XP_754048); Afu2g05330 (vacuolar H+/Ca2+ exchanger: XP_749663); Afu1g11110 (calcium channel subunit Cch1: XP_752476); Afu4g04670 (calcium permease family membrane transporter: XP_746653); Afu2g07630 (vacuolar H+/Ca2+ exchanger: XP_755098); Afu2g05320 (calcium-proton exchanger: XP_749662); Afu1g04270 (calcium ion transporter Vcx1: XP_750174); and Afu4g03320 (similar to vacuolar H + /Ca 2+ exchanger: XP_001481534). doi:10.1371/journal.pone.0037591.g001 MM+25 mM MnCl 2 ( Figure 6C). Both the DpmcA::pmcA + and DpmcB::pmcB + showed the same phenotype as the wild-type strain, strongly indicating that the null phenotypes observed for both genes were only due to the introduction of these mutations in the corresponding strains ( Figures 5 and 6).
Since the DpmcA and DpmcB strains were calcium-sensitive but manganese-resistant, we decide to investigate the mRNA abundance of each pmc gene when the wild-type, DpmcA and DpmcB mutant strains were exposed to a short pulse of MnCl 2 ( Figure 7). All three genes showed increased mRNA abundance in the presence of MnCl 2 (pmcA and pmcB have 2.5-and 2.0-fold more transcripts after 10 minutes; Figures 7A and B), however the highest induction was observed for pmcC that showed a 30-and 3.7-fold increase in transcripts after 10 and 30 minutes, respectively ( Figure 7C). Nevertheless, like observed for calcium induction, the absolute mRNA levels of pmcC are the lowest among all the three genes ( Figure 7). The pmcA mRNA levels in the DpmcB mutant strain exposed to MnCl 2 were about the same as the wild-type strain ( Figure 7A). When the DpmcA mutant strain was exposed to MnCl 2 , the mRNA levels of pmcB were 2.5-fold higher than the wild-type strain after 10 minutes exposure. Interestingly, the pmcB mRNA levels in this mutant without any MnCl 2 exposure (i.e., the control before exposure) were 2.3-fold higher than the wild-type strain ( Figure 7B). Finally, there was a decrease in the pmcC mRNA levels after DpmcA and DpmcB mutant strains were exposed to MnCl 2 ( Figure 7C). Down-regulation or overexpression of pmcC had no effect on pmcA or pmcB mRNA accumulation (data nor shown).
Finally, we evaluated the relative concentration of free calcium in the A. fumigatus wild-type, DpmcA, DpmcB, DpmcA::pmcA + , and DpmcB::pmcB + strains by using Fura-2-AM, a highly sensitive dye for rapid measurement of calcium flux in cells (www.invitrogen. com). Fura-2-Am is a fluorescent calcium indicator that can passively diffuse across cell membranes and when inside the cell, the esters are cleaved by intracellular esterases to yield cellimpermeant fluorescent indicator. Upon binding Ca +2 , Fura-2 exhibits an absorption shift from 380 to 340 nm of excitation. Thus, the relative Ca +2 concentration was evaluated based on the fluorescence ratio after dual-wavelength excitation. Upon calcium exposure, the DpmcA mutant strain had an increased relative level of intracellular calcium concentration compared to the same strain in the absence of calcium ( Figure 8). This difference is not observed for the wild-type, DpmcB and complemented strains, and alcA::pmcC strain (data not shown).
We have not observed any differential susceptibility of these mutants to antifungal agents, such as amphotericin, azoles, and caspofungin, in E-tests (data not shown). These results indicate pmcA and pmcB are involved in calcium and manganese metabolism in A. fumigatus, and also suggest pmcA is the major transporter responsible for removing calcium from the cytoplasm. Expression of the pmcA-C genes in murine-infecting A. fumigatus wild-type, DcalA and DcrzA mutant strains Previously, we generated by RNA amplification multiple gene expression profiles via minute samplings of A. fumigatus germlings during the initiation of murine infection [25,26]. This enabled us to identify genes preferentially expressed during adaptation to the mammalian host niche. Here, we took advantage of the establishment of this technical platform to characterize genes that have in vivo decreased or increased mRNA abundance in the DcalA and DczA mutant strains when compared to the wild-type strain.
We firstly characterized the time course of hyphal development in the sequenced clinical isolate Af293, DcalA and DcrzA mutant strains by histopathological examination of infected neutropenic murine lung tissues (Supplementary Figure S4). Lung sections collected and formalin-fixed at 4, 10 and 14 hours post-infection contained numerous A. fumigatus wild-type, DcalA and DcrzA spores in close association with murine epithelium in the bronchioles and   Figure S4, upper panels). At 12-14 hours post-infection, 80% of A. fumigatus conidia from the three strains had undergone comparable germination and primary hyphal production. Bronchoalveolar lavage was performed immediately using pre-warmed sterile saline and samples (BALFs) were snap frozen prior to RNA extraction and amplification. Within infection groups BALFs were pooled prior to RNA extraction and mRNA amplification. Total RNA extracted from these cultures was used to amplify fluorescent-labeled cDNAs for real-time PCR experiments. We designed Lux fluorescent probes and used realtime RT-PCR analysis to quantify the pmcA, pmcB, and pmcC mRNA abundance in the DcalA and DcrzA germlings after bronchoalveolar lavage at 4 and 14 hours growth and compared this with their expression in the wild-type strain grown during the  same time points (reference treatment) ( Table 1). All these three genes showed, to different extents, decreased mRNA abundance during the initiation of murine infection by the A. fumigatus DcalA and DcrzA mutant strains relative to the wild-type strain. Thus, it seems that in vivo pmcA-C mRNA accumulation is dependent on CalA and crzA.
The A. fumigatus DpmcA mutant strain is avirulent in low dose murine infection To assess the role of PmcA-B in pathogenicity we tested the A. fumigatus DpmcA-B mutant strains in a neutropenic murine model of invasive pulmonary aspergillosis, comparing virulence of the A. fumigatus DpmcA-B mutant strains (n = 10 for each mutant) to that of the wild-type (n = 10) ( Figure 9A and Supplementary Figure S5). While infection with the wild-type strain resulted in a mortality rate of over 100% at 6 days post-infection, infection with the pmcA deletion strain resulted in a significantly reduced mortality rate of approximately 20% after 10 days post-infection (p,0.005). The pmcB mutant showed virulence comparable to the wild-type strain (Supplementary Figure S5). Since the comparison between DpmcA infected group and the non-infected group (PBS) showed to be statistically non-significant (p = 0.1451), we can consider this strain avirulent. To directly link the observed attenuated virulence of the DpmcA mutant with the replacement of the DpmcA locus we tested an independent strain resulting from single ectopic reintegration of the wild-type DpmcA locus (Supplementary Figure S6) and with the complementation strain full virulence was restored ( Figure 9A). To further understand the basis of attenuated virulence in the DpmcA background we made histopathological examinations of infected tissues at early time points in infection, aiming to identify differences in growth rate, tissue invasion and inflammatory responses between the two strains. At 72 hours post-infection the lungs of mice infected with the wild-type isolate contained multiple foci of invasive hyphal growth, manifesting as both penetration of the pulmonary epithelium in major airways ( Figure 9B) and pockets of branched invading mycelia originating from the alveoli ( Figure 9B). In contrast, infection resulting from DpmcA inoculations was typified by contained inflammatory infiltrates in bronchioles ( Figure 9B) some of which contained fungal elements in the form of poorly germinated or ungerminated spores. Fungal burden data as measured by real-time PCR showed that the DpmcA mutant strain did not grow within the lungs as well as the wild-type and the complemented DpmcA strains ( Figure 9C, p,0.0001). Taken together, these data strongly indicate that PmcA plays a role in A. fumigatus virulence.

Discussion
We have been actively looking for additional components of the Ca +2 -calcineurin pathway [14,17,27]. One of these components, the transcription factor CrzA induces the expression of various cation transporters that act at the plasma membrane or on other membranous organelles [14,17]. Very little is known about calcium transport and calcium homeostasis in filamentous fungi. Most of our knowledge about calcium homeostasis in fungi is derived from S. cerevisiae, where more than 95% of cellular calcium is sequestered in the vacuole [28,29,30]. In S. cerevisiae PMC1, PMR1, and VCX1 encode a vacuolar Ca 2+ ATPase involved in depleting cytosol of Ca 2+ ions, a high affinity Ca 2+ /Mn 2+ P-type ATPase required for Ca 2+ and Mn 2+ transport into the Golgi, and a vacuolar membrane antiporter with Ca 2+ /H + and K + /H + exchange activity, involved in the control of cytosolic Ca 2+ and K + concentrations, respectively [19,22,31]. S. cerevisiae PMC1 knockout mutants sequester Ca +2 into the vacuole at 20% of the wildtype levels and fail to grow in media containing high levels of Ca +2 [19]. Mutations in the calcineurin A or B subunits or the addition of FK506 or cyclosporin A restored growth of pmc1 mutants in media with high Ca +2 concentrations [20,21]. In Neurospora crassa it was reported that active transport across the plasma membrane is important for keeping low levels of cytosolic calcium [32,33]. In addition, in this species the vacuole is important for regulating the intracellular calcium levels [32,34,35]. In N. crassa, there are two PMC1 homologues, NCA-2 and NCA-3 [36]. The NCA-2 fused with GFP is located in the plasma membrane as well as in vacuolar membranes in this organism [37], suggesting NCA-2 functions to pump calcium out of the cell. The Dnca-3 strain showed comparable levels of calcium sensitivity to the wild-type strain; in contrast, Dnca-2 showed significant inhibition of growth at 50 mM CaCl 2 and accumulates 10-fold more intracellular calcium than the wild-type strain [36]. In A. nidulans, two null mutants constructed for the PMC1 homologues, pmcA and pmcB, displayed low-sensitivity to 700 mM CaCl 2 concentrations [38]. However, the double A nidulans DpmcA DpmcB mutant has increased calciumsensitivity suggesting these two genes are genetically interacting [38].
Rispail et al. [39] have proposed that A. fumigatus has three PMC1, two PMR1, and four VCX1 homologues. Here, we have concentrated our attention on three genes encoding PMC1 calcium transporter homologues that have their mRNA levels dependent on CalA-CrzA [14,17], generated mutants of them, and studied their phenotypes and virulence. Although A. fumigatus pmcA-C genes are involved in calcium metabolism, this work did not provide a full characterization of their function. We do not know their sub-cellular localization and how they affect subcellular calcium abundance. We were not able to knock-out pmcC and subsequently demonstrated that pmcC is an essential A. fumigatus gene by constructing a conditional pmcC mutant. The pmcC downregulation causes growth inhibition and its overexpression can produce a physiological imbalance, as the mutant strain also has reduced growth. We were able to demonstrate that CrzA can control the pmcA-C mRNA expression by binding directly to their promoter regions. Crz1p has a C2H2 zinc finger motif that binds to CDRE in the promoters of genes that are regulated by calcineurin and calcium (Stathopoulos and Cyert, 1997). Yoshimoto et al. Here, we demonstrated that CrzA can bind directly to 300-bp upstream regions from pmcA-C genes. In two of these genes, pmcA and pmcC, we were able to identify putative CDRE motifs and demonstrated that they can completely inhibited the complexes formed with pmcA and pmcC DNA fragments. These results strongly suggest these CDRE motifs are functional and this is probably the first demonstration of CDRE functionality in a human pathogenic fungus.
Cyclosporin was able to modulate the sensitivity or resistance of these mutants to either calcium or manganese chloride, once more supporting the interaction between calcineurin and the function of these transporters. In addition, we showed wild-type levels of susceptibility to amphotericin B, voriconazole, posoconazole, itraconazole, and caspofungin (E-test assays) and that there were no defects in cell wall integrity (data not shown). We also observed that the complete and minimal culture media affected the susceptibility of the DpmcA and DpmcB mutant strains to calcium and manganese chloride. The defined macronutrients composition of MM medium could explain the differences in growth of the DpmcA and DpmcB mutant strains in YAG and MM media. The MM is composed of glucose, trace elements, and macronutrients (salt solution). The salt solution is composed of sodium nitrate, potassium chloride, potassium phosphate, and magnesium sulphate. When the wild-type, DpmcA, and DpmcB strains are grown in MM supplemented only with a single one of these macronutrients, there is a reduction in radial growth for all strains, except for MM+MgSO 4 that showed about the same radial growth as in MM (data not shown). The most likely reasons for this outcome are either the mechanism of action of these transporters depends on other cations, such as sodium, or there is some cross-talk with the mechanisms for ion detoxification. Recently, Spielvogel et al. [42] have shown that SltA, a transcription factor important for cation adaptation and homeostasis acts positively on the transcription of the Ena1p-like Na + pump gene enaA and negatively on the transcription of the putative vacuolar Ca +2 /H + exchanger gene vcxA (A. fumigatus homologue is Afu1g04270). Interestingly, the negative regulation of vcxA by SltA is opposed by its transcriptional activation by CrzA [42].
A. fumigatus pmcA-C genes have decreased mRNA abundance into the alveoli in the DcalA and DcrzA mutant strains. Accordingly, when A. fumigatus is exposed in vitro to calcium chloride, there is a decrease in pmcA-C mRNA abundance in both mutants. When we compare the absolute pmcA-C mRNA abundance levels in the wildtype strain grown in mouse alveoli, we observed that pmcB has about five to ten times higher levels than pmcA, while pmcC has very low levels of mRNA abundance (1,000 to 3,000 times lower than pmcB). Consistently, the same mRNA abundance is observed when A. fumigatus is exposed in vitro to calcium chloride. Interestingly, there is an increase in the pmcB mRNA levels in the DpmcA mutant strain when this strain is not exposed to CaCl 2 , suggesting a compensation for the pmcA absence. An intriguing observation from our work is the fact that pmcC has very low absolute levels of mRNA accumulation in all conditions tested in this work, but it is an essential gene. This is confirmed by a weak CrzA binding to pmcC promoter. It is possible that PmcC specific activity is very high and this will compensate its low mRNA levels. It is also possible that PmcC has other functions that were not identified in this work and are essential for cell metabolism. Interestingly, both DpmcA and DpmcB mutants are more resistant to MnCl 2 than the wild-type strain and had reduced pmcC mRNA accumulation when exposed to either CaCl 2 or MnCl 2 . These results suggest pmcC mRNA levels are dependent on pmcA and pmcB, when A. fumigatus is exposed either to calcium or manganese. However, this effect is more notable in the presence of manganese.
The DpmcA mutant is avirulent in a neutropenic murine model of invasive pulmonary aspergillosis. The reduced virulence of the DpmcA could be due to an excess of calcium in the cytoplasm that could not be removed due to the lack of pmcA, thus potentially affecting several functions such as secretion, cell wall composition and the activation of pathways necessary for infection. Interestingly, we did not observe attenuated virulence for DpmcB, suggesting that the different PMC1 paralogues have different functions during pathogenicity. This is the first demonstration of the involvement of a calcium transporter in A. fumigatus virulence. Previously, Pinchai et al. [16] have shown that A.fumigatus DpmrA has several defects related to growth, cationic tolerance, and increased beta-glucan and chitin content, but in spite of all these abnormal phenotypes the mutant strain remained virulent.
In conclusion, we have shown that PmcA is required for full virulence in animal infection. In addition, that PmcA acts in the A. fumigatus Ca +2 -calcineurin signaling pathway and influences the relative intracellular calcium concentration. Further studies are necessary to address the sub-cellular location of PmcA, -B, and -C, and how PmcA contributes to the pathogenesis of aspergillosis.

Ethics statement
The principles that guide our studies are based on the Declaration of Animal Rights ratified by the UNESCO in January 27, 1978 in its articles 8 th and 14 th . All protocols used in this study were approved by the local ethics committee for animal experiments from the Campus of Ribeirão Preto from Universidade de Sao Paulo (Permit Number: 08.1.1277.53.6; studies on the interaction of Aspergillus fumigatus with animals). All animals used in this study were housed in groups of five in individually ventilated cages and were cared for in strict accordance to the principles outlined in the by the Brazilian College of Animal Experimentation (Princípios É ticos na Experimentação Animal -Colégio Brasileiro de Experimentação Animal, 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 by a veterinarian and humanely sacrificed if moribund (defined by lethargy, dyspnoea, hypothermia and weight loss).   variants: YAG (2% w/v glucose, 0.5% w/v yeast extract, 2% w/v agar, trace elements), YUU [YAG supplemented with 1.2 g l-1 (each) of uracil and uridine], and liquid YG or YG + UU medium with the same composition (but without agar). A modified minimal medium (MM: 1% glucose, original high nitrate salts, trace elements, 2% agar, pH 6.5) was also used. Expression of pmcC gene, under the control of alcA promoter, was regulated by carbon source: repression on glucose 4% w/v, derepression on glycerol, and induction on threonine. Therefore, MM + Glycerol and MM + Threonine were identical to MM, except that glycerol (2% v/v) and/or threonine (100 mM) were added, respectively, in place of glucose as the sole carbon source. Trace elements, vitamins, and nitrate salts were included as described by [43]. Strains were grown at 37uC unless indicated otherwise. Additionally, 10% fetal bovine serum (Gibco) was used as a medium.

Construction of the A. fumigatus mutants
A gene replacement cassette was constructed by ''in vivo'' recombination in S. cerevisiae as previously described [44]. Briefly, approximately 2.0 kb regions on either side of each ORF were selected for primer design. For construction, the primers were named as 5F and 5R, were used to amplify the 59-UTR flanking region of the targeted ORF. Likewise, the primers 3F and 3R were used to amplify the 39-UTR ORF flanking region, and the primers 5F and 3R also contains a short homologue sequence to the MCS of the plasmid pRS426. Both fragments, 5-and 3-UTR, were PCR-amplified from A. fumigatus genomic DNA (gDNA). The pyrG used in the A. fumigatus cassette for generating the mutant strains were used as marker for prototrophy. Deletion cassette generation was achieved by transforming each fragment along with the plasmid pRS426 BamHI/EcoRI cut in the in S. cerevisiae strain SC94721 by the lithium acetate method [45]. The DNA of the yeast transformants was extracted by the method described by Goldman et al. [46], dialysed and transformed by electroporation in Escherichia coli strain DH10B to rescue the pRS426 plasmid harboring the cassette. The cassette was PCR-amplified from these plasmids and used for A. fumigatus transformation. Southern blot analyses were used throughout of the manuscript to demonstrate that the transformation cassettes had integrated homologously at the targeted A. fumigatus loci. For the construction of the alcA::pmcC strain, 1000 bp of the pmcC encoding region was cloned downstream to the alcA promoter into the pMCB17apx vector. This construction was further transformed in A. fumigatus to replace the endogenous pmcC promoter yielding the strain alcA::pmcC. The DpmcA mutant strain was complemented by co-transforming a pmcA + DNA fragment (approximately 1 kb from each 59 and 39flanking regions plus the ORF) together with the pHATa vector [47] and selecting for hygromycin resistance in MM plates with 150 mg/ml of hygromycin B.

RNA extraction and real-time PCR reactions
After treatment conditions, mycelia were harvested by filtration, washed twice with H 2 O and immediately frozen in liquid nitrogen. For total RNA isolation, the germlings were disrupted by grinding in liquid nitrogen with pestle and mortar. Total RNA was extracted with Trizol reagent (Invitrogen, USA). Ten micrograms of RNA from each treatment was then fractionated in 2.2 M formaldehyde, 1.2% w/v agarose gel, stained with ethidium bromide, and then visualized under UV light. The presence of intact 25S and 18S ribosomal RNA bands was used to assess the integrity of the RNA. RNasefree DNase I treatment, for the realtime PCR experiments, was carried out as previously described [48]. Twenty micrograms of total RNA was treated with DNase, purified using a RNAeasy kit (Qiagen) and cDNA was generated using the SuperScript III First Strand Synthesis system (Invitrogen) with oligo(dT) primers, according to the manufacturer's protocol.
All the PCR reactions were performed using an ABI 7500 Fast Real-Time PCR System (Applied Biosystems, USA) and Taq-Man Universal PCR Master Mix kit (Applied Biosystems, USA). The reactions and calculations were performed according to Semighini et al. [48]. The primers and Lux TM fluorescent probes (Invitrogen, USA) used in this work are described in Supplementary Table S1.

Cloning the crzA gene into the pDEST15 vector
The Gateway Technology (Invitrogen) was used to construct, in Escherichia coli, the expression system consisting of the CRZA gene N-tagged to the GST gene. Briefly, the coding region of the exon 2 from CrzA was amplified from the cDNA sample by PCR using PlatinumH Taq DNA Polymerase High Fidelity. (Invitrogen) and specific primers (CRZ-exon2-attB1-F 59-GGGGACAAGTTG-TACAAAAAAGCAGGCTTCGAAGGAGATAGAAC-CATGTCCCGCGGGCGTAGCAAG-39 and CRZ-attB2-R 59-GGGGACCACTTTGTACAAGAAAGCTGGGTCTCAATA-GAAGTTACCGGCAGCAG-39). Amplification was run for 30 cycles consisting of denaturation at 94uC for 1 min, primer annealing at 55uC for 1 min and primer extension at 68uC for 2 min. The PCR product carrying the attB sites was purified from agarose gel using the QIAquick PCR purification kit (Qiagen) and cloned into the pDONR201 plasmid (Invitrogen) using the BP Clonase. The BP clonase catalyze the in vitro recombination of PCR products or DNA segments from clones (containing attB sites) and a donor vector (containing attP sites) to generate entry clones. The entry clone pDONR201-CrzA was transformed into E. coli DH10B competent cells and selected for kanamicyn resistence. Entry clones were checked by sequencing using the ATT primers (ATT1F-59-TCGCGTTAACGCTAGCATG-GATCTC-39 and ATT2 R-59-GTAACATCAGAGATTTT-GAGACAC-39) and further used in LR reactions (Invitrogen) with the vector pDEST15 (an N-terminal GST fusion vector containing the T7 promoter) to generate the expression vector pDEST15-GST/CrzA.

Production and Purification of GST::CrzA
A. nidulans CrzA was expressed as a GST-fusion protein from the construct pDEST15-GST/CrzA in E. coli Rosetta TM (DE3) pLysS strain (Novagen). Cells harboring the plasmid construction were grown in 1 L of LB medium to an O.D. 600 nm of 0.8 and protein expression was induced at 12uC, 180 rpm overnight with 0.4 mM IPTG final concentration. After induction, cells were harvested by centrifugation, suspended in phosphate-buffered saline solution (500 mM NaCl, 2.7 mM KCl, 100 mM Na 2 HPO 4 , 2 mM KH2PO4, 5% v/v glycerol, 0.5% NP-40, pH 7.4) containing 10 mM benzamidine, 0.5 mM EDTA and 2 mM of each DTT and PMSF and lysed by sonication (ten 30 sec pulses on ice) in a Vibra-Cell disrupter (SonicsH). Cell lysate was clarified at 23,0006 g, 20 min, 4uC, and the recombinant protein was purified by affinity chromatography on a GSTrap FF column (GE HealthCare) according to manufacturer's instructions on an Ä KTA Prime purification system. Recombinant protein was eluted in a linear gradient of 20 mM glutathione in 50 mM Tris-HCl, 500 mM NaCl, 5% v/v glycerol, 2 mM DTT, pH 8.0 buffer. Chromatographic fractions were analyzed by SDS-PAGE followed by Coomassie Brilliant Blue staining [49] and fractions containing the purified protein were combined, concentrated and quantified using BSA as standard [50].

Electrophoretic Mobility Shift Assay
GST::CrzA recombinant protein was assayed in DNA-protein binding reactions using three 300 bp DNA fragments of the pmcA, pmcB and pmcC promoters as probes, containing the putative cisregulatory calcineurin-dependent response elements (CDREs) for the transcription factor CrzA (Supplementary Table S1 and Figure  S1). Binding reactions were carried out in 16binding buffer (25 mM HEPES-KOH, pH 7.9, 20 mM KCl, 10% w/v glycerol, 1 mM DTT, 0.2 mM EDTA, pH 8.0, 0.5 mM PMSF, 12.5 mM benzamidine, 5 mg/mL of each antipain and pepstatin A) containing 2 mg poly(dI-dC).(dI-dC) as non-specific competitor and 1-2 mg of GST::CrzA recombinant protein, at room temperature for 10 min. After that, DNA probes (10 4 cpm) were added and the binding reactions were incubated at room temperature during 20 min prior to being loaded onto a native 5% polyacrylamide gel in 0.56 TBE buffer. Gels were run at 10 mA, 15uC, dried, and exposed to X-ray film. For competition assays, a molar excess of the specific DNA competitors were added prior to incubation with the radiolabeled probe.

DNA Probes and Specific Competitors for EMSA
Putative cis CrzA motifs were visually identified in the promoter regions of the genes pmcA, pmcB and pmcC by using the A. fumigatus CDRE consensus. To produce the pmcA probe, a 300 bp DNA fragment of the pmcA promoter was amplified from A. fumigatus genomic DNA by using the primers PMCA-5R and 59-PMCA (Supplementary Table S1) in the presence of [a-32 P]-dATP (3,000 Ci/mmol) and purified on 2% low-melting point agarose gel. pmcB and pmcC probes were prepared as above using the primer pairs PMCB-5R and 59-PMCB (Supplementary Table S1), and PMCC-5R and 59-PMCC (Supplementary Table S1). The unlabeled 300 bp pmcA, pmcB and pmcC probes were used as specific DNA competitors which were quantified by measuring the absorbance at 260 nm and added to the binding reaction in a 30to 50-fold molar excess, 10 min prior to the addition of the respective probes. DNA oligonucleotides containing the CDRE motifs identified in pmcA and pmcC probes were also used as specific competitors after annealing the complementary oligonucleotides pairs pmcA1/pmcA2 and pmcC1/pmcC2, respectively (Supplementary Table S1). The DNA oligonucleotides were quantified by measuring the absorbance at 260 nm and added to the reaction in 10-30 fold molar excess.
Mutated probes (mpmcA and mpmcC) were prepared by changing the element core sequences by site-directed mutagenesis in a two-step PCR. In the mpmcA probe the sequence 59-CCCTGCCCC-39 was changed to 59-AAAGTAAAA-39 by using the oligonucleotide pair mPMCA-F and mPMCA-R in the in the first reaction to amplify two fragments. The oligonucleotide pair PMCA-5R and 59-PMCA (Supplementary Table S1) was used in a second reaction to amplify the whole DNA fragment containing the mutation. In the mpmcC probe the sequence 59-CACAGC-CAC-39 was changed to 59-ACACTAACA-39 by using the oligonucleotide pair mPMCC-F and mPMCC-R in the in the first reaction. The oligonucleotide pair PMCC-5R and 59-PMCC (Supplementary Table S1) was used in a second reaction to amplify the whole DNA fragment containing the mutation. For EMSA, both mutated fragments were used as templates in PCR amplifications in the presence of [a-32 P]-dATP (3,000 Ci/mmol) and purified on 2% low-melting point agarose gel.

Determination of the relative levels of intracellular calcium concentration
To investigate the relative intracellular free calcium concentration we used the Fura-2 acetoxymethyl ester (Fura-2-AM; Invitrogen). Briefly, 10 7 conidia of each wild-type, DpmcA, DpmcA::pmcA + , DpmcB, and DpmcB::pmcB + were incubated in YG medium for 8 hours with shaking at 37uC. Then, each strain was either treated with 500 mM CaCl 2 , or not, in fresh YG medium for 30 minutes. After incubation the cells were washed three times with PBS and loaded with 10 mM Fura-2-AM for 30 min at 37uC. After washing, Fura-2 fluorescence was measured by alternating the excitation wavelengths at 340 and 380 nm with an emission wavelength fixed at 505 nm. The relative intracellular calcium concentration is expressed as the ratio between fluorescence intensities with excitation wavelengths at 340 and 380 nm. All data presented are representative of three independent experiments.

Murine model of pulmonary aspergillosis
Outbred female mice (BALB/c strain, 20-22 g) were housed in individually vented cages, containing 5 animals. Mice were immunosuppressed with cyclophosphamide at 150 mg/kg of body weight, administered intraperitoneally on days 24, 21 and 2, and hydrocortisonacetate was injected subcutaneously at 200 mg/kg on day 23, modified from [51]. A. fumigatus spores for inoculation were grown on Aspergillus complete medium for 2 days prior to infection. Conidia were freshly harvested using sterile PBS and filtered through Miracloth (Calbiochem). Conidial suspensions were spun for 5 min at 3000 g, washed three times with sterile PBS, counted using a hemocytometer and re-suspended at a concentration of 2.5610 6 conidia/ml. Viable counts from administered inocula were determined following serial dilution by plating on Aspergillus complete medium and grown at 37uC. Mice were anaesthetized by halothane inhalation and infected by intranasal instillation of 5.0610 4 conidia in 20 ml of PBS. As negative control, a group of 5 mice received only PBS intranasally. Mice were weighed every 24 h from the day of infection and visually inspected twice daily. In the majority of cases the endpoint for survival experimentation was when a 20% reduction in body weight measured from the day of infection and at this point the mice were sacrificed. Significance of comparative survival was calculated using Log Rank analysis in the Prism statistical analysis package. Additionally, at 3 days post infection, 2 mice per strain were sacrificed, from which the lungs were removed, fixed and processed for histological analysis.

Lung histopathology and fungal burden
After sacrifice, the lungs were removed and fixed for 24 h in 10% buffered formalin phosphate. Samples were washed in 70% alcohol several times, dehydrated in alcohols of increasing concentrations, diafanized in xylol and embedded in paraffin. For each sample, sequential 5 mm sections were collected on glass slides and the sections were stained with Gomori methenamine silver (GMS) or hematoxylin and eosin (HE) stain following standard protocols [24]. Briefly, sections were deparaffinized, oxidized with 4% chromic acid, stained with methenamine silver solution, and counter stained with picric acid or light green. For HE staining, sections were deparaffinized, stained first with hematoxylin and then stained with eosin. All stained slides were immediately washed, preserved with mounting medium and sealed with a cover glass. Microscopical analyses were done using an Axioplan 2 imaging microscope (Zeiss) at the stated magnifications under brightfield conditions.
To investigate fungal burden in murine lungs, mice were immunosuppressed with cyclophosphamide at 150 mg/kg of body weight administred intraperitoneally on days 24 and 21 and hydrocortisonacetat injected subcutaneously at 200 mg/kg on day 23. Five mice per group (wild-type, DpmcA, DpmcA::pmcA, and PBS control) were inoculated with 5610 5 conidia/20 ml suspension intranasally. A higher inoculum, in comparison to the survival experiments, was used to increase fungal DNA detection. Animals were sacrificed 48 hours post infection, both lungs were harvested and immediately frozen in liquid nitrogen. A mortar and pestle were used to pulverize the samples (frozen in liquid nitrogen) and DNA was extracted by the Phenol/Chlroform method. DNA quantity and quality was assessed with a NanoDrop 2000 (Thermo Scientific). Around 200 ng of total DNA of each sample was used for quantitative Real-Time PCR reaction. A primer and a Lux TM probe (invitrogen) were used to amplify the 18S rRNA region of A. fumigatus (primer: 59-CTTAAATAGCCCGGTCCGCATT-39, probe: 59-CATCACAGACCTGT TATTGCCG-39) and an intronic region of mouse GAPDH (primer: 59-CGAGG-GACTTGGAGGACACAG-39, probe: 59-GGGCAAGGC-TAAAGGTCAGCG-39). Six-point standard curves were calculated using serial dilutions of gDNA from all A. fumigatus strains used here and non-infected mouse lung. Fungal and mouse DNA quantities were obtained from the Ct values from an appropriate standard curve. Fungal burden was determined via the ratio between ng of fungal and mouse DNA.

Bronchoalveolar lavages
To analyze gene expression of A. fumigatus strains during early pulmonary infection, Outbred female mice (BALB/c strain, 20-22 g) were housed in individually vented cages, containing 5 animals. Mice were immunosuppressed with cyclophosphamide (Genuxal, Baxter) at 150 mg/kg of body weight administered intraperitoneally on days 24 and 21 and hydrocortisone sodium succinate (Hidrosone, Cellofarm) was injected subcutaneously at 200 mg/kg on day 21. All mice received tetracycline hydrochloride 0.5 mg/L in drinking water, as prophylaxis against bacterial infection. A. fumigatus spores for inoculation were grown on Aspergillus complete solid medium (YAG) for 2 days prior to infection. Conidia were freshly harvested using sterile PBS and filtered through Miracloth (Calbiochem). Conidial suspensions were spun for 5 min at 4,000 rpm, washed three times with sterile PBS, counted using a hemocytometer and re-suspended at a concentration of 2.5610 10 conidia/ml. Five mice per group (wildtype, DcrzA and DcalA) were anesthetized by isoflurane (Isothane, Baxter) inhalation and infected by intranasal instillation of 10 9 conidia in 40 ml of PBS. Groups of infected mice were sacrificed and processed collectively at time points 4 and 12 hours postinfection. Bronchoalveolar lavage (BAL) was performed immediately after culling using three 0.5 ml aliquots of cold sterile PBS. To remove the mice cells from BALs, samples were spun down in microcentrifuge tubes, supernatants were removed, the samples were resuspended in 1 ml of sterile ultrapure water, centrifuged again and finally the pellets were snap frozen immediately using liquid nitrogen. To extract RNA, BAL samples from each strain (5 BALs per strain) were mixed with 1 ml Trizol LS Reagent (Invitrogen) and acid treated glass beads (425-600 mm, Sigma-Aldrich). Fungal cells were homogenized by 10 min vortexing, centrifuged at 12,000 rpm for 10 min, the upper phase was mixed with 200 ml chloroform, centrifuged again, the new upper phase was mixed with 500 ml isopropanol and incubated overnight at 280uC. After washing the pellet with 70% ethanol, RNA was dissolved in 20 ml DEPC water. Further RNA purification was carried out using RNeasy mini kit (Qiagen), following manufacturer's instructions. RNA concentration and integrity was monitored by NanoDropH 2000 -Thermo Scientific (Uniscience). RNA amplification was done according to Agilent Low RNA Input Fluorescent Linear Amplification kit (Agilent Technologies). RNAse free DNAse treatment was carried out as previously described [48].