The metalloreductase FreB is involved in adaptation of Aspergillus fumigatus to iron starvation.

Aspergillus fumigatus employs two high affinity iron uptake mechanisms, siderophore mediated iron uptake and reductive iron assimilation (RIA). The A. fumigatus genome encodes 15 putative metalloreductases (MR) but the ferrireductases involved in RIA remained elusive so far. Expression of the MR FreB was found to be transcriptionally repressed by iron via SreA, a repressor of iron acquisition during iron sufficiency, indicating a role in iron metabolism. FreB-inactivation by gene deletion was phenotypically largely inconspicuous unless combined with inactivation of the siderophore system, which then decreased growth rate, surface ferrireductase activity and oxidative stress resistance during iron starvation. This study also revealed that loss of copper-independent siderophore-mediated iron uptake increases sensitivity of A. fumigatus to copper starvation due to copper-dependence of RIA.

Introduction

Iron is an indispensable trace element for all eukaryotes and almost all prokaryotes. As a transition element iron can adopt two ionic forms, the reduced ferrous (Fe2+) and the oxidized ferric (Fe3+) state. The capacity to accept or donate electrons makes iron the major redox mediator in biology. Either alone or incorporated into iron–sulfur clusters or heme, iron is required for fundamental cellular processes like the tricarboxylic cycle, oxidative stress detoxification as well as biosynthesis of amino acids, desoxyribonucleotides, and sterols. However, iron excess has the ability to generate toxic reactive species that can damage cellular components (Halliwell and Gutteridge, 1984). Despite its high abundance in the Earth’ crust, the bioavailability of iron is very limited owing to its oxidation into insoluble Fe3+-hydroxides by atmospheric oxygen. Consequently, all organisms have developed tightly regulated mechanisms in order to balance uptake, storage and consumption of iron.

Aspergillus fumigatus is a typical ubiquitous saprophytic mold. Nevertheless, it causes life-threatening invasive diseases especially in immuno-compromised patients and has become the most common airborne fungal pathogen of humans (Tekaia and Latge, 2005). A. fumigatus employs two high-affinity iron uptake systems, siderophore-assisted iron uptake and reductive iron assimilation (RIA), both of which are induced upon iron starvation (Schrettl et al., 2004, 2007, 2008). Siderophores are low molecular mass, Fe3+-specific chelators (Haas et al., 2008, 2003). A. fumigatus excretes the siderophores fusarinine C and triacetylfusarinine C (TAFC) to mobilize extracellular iron. Subsequent to chelation of iron, the Fe3+-forms of fusarinine C and TAFC are incorporated by specific transporters, e.g. MirB (Haas et al., 2003). A. fumigatus possesses also intracellular siderophores, hyphal ferricrocin and conidial hydroxyferricrocin, for distribution and storage of iron (Schrettl et al., 2007; Wallner et al., 2009). RIA starts with reduction of Fe3+-sources to the more soluble Fe2+ by plasma membrane-localized ferrireductases, which have not been identified in A. fumigatus yet (Kosman, 2010). Subsequently, Fe2+ is re-oxidized and imported by a protein complex consisting of the ferroxidase FetC and the iron permease FtrA. Ferroxidases belong to the multicopper oxidase family resulting in copper-dependence of RIA (Hassett et al., 1998; Kosman, 2010). Moreover, A. fumigatus employs low affinity iron uptake, which has not been characterized at the molecular level (Schrettl et al., 2004).

Both extra- and intracellular siderophores contribute to pathogenic growth because elimination of the entire siderophore system (ΔsidA mutant) results in absolute avirulence of A. fumigatus in a murine model of invasive pulmonary aspergillosis, while deficiency in either extracellular (ΔsidF or ΔsidD mutants) or intracellular siderophores (ΔsidC mutants) causes partial attenuation of virulence (Schrettl et al., 2004, 2007). Genetic inactivation of RIA (ΔftrA mutant) does not affect virulence of A. fumigatus (Schrettl et al., 2004, 2007). Nevertheless, several lines of evidence indicate that RIA also plays a role during infection: (i) elimination of extracellular siderophores causes only partial attenuation of virulence, (ii) genome-wide expression profiling revealed induction of both the siderophore system and RIA during initiation of murine infection, and (iii) mutants lacking both RIA and the siderophore system (ΔftrAΔsidA double mutant) are unable to grow unless supplemented with siderophores or extremely high iron concentrations fueling low-affinity iron uptake (McDonagh et al., 2008; Schrettl et al., 2004, 2007). Of note, RIA has been shown to be crucial for virulence in animal models of the siderophore-lacking species Candida albicans and Cryptococcus neoformans and the siderophore-producing species Rhizopus oryzae (Eichhorn et al., 2006; Ibrahim et al., 2010; Jung et al., 2008; Ramanan and Wang, 2000). Moreover, RIA is essential for virulence of the plant-pathogenic siderophore-producing species Ustilago maydis (Eichhorn et al., 2006; Jung et al., 2008; Ramanan and Wang, 2000).

Metalloreductases (MR) belong to the integral membrane reductases (IMR) protein family (Grissa et al., 2010). IMR are present in all fungal species and can be classified into at least 24 subgroups. IMR are homologous to fungal NADPH oxidases (NOX), which reduce oxygen to superoxide () for antimicrobial defense (e.g. mammalian gp91phox) or signaling (Aguirre et al., 2005; Frey et al., 2009). IMR activity requires NADPH, FMN, and heme. Like NOX, IMRs oxidize cytoplasmic NADPH and transfer the electron across the plasma membrane to reduce small molecules, dioxygen in the case of NOX (yielding thus the superoxide anion) and mostly metals (Fe3+ and/or Cu2+) in the case of IMRs. Fungal MR have so far been biochemically or genetically characterized exclusively in three yeast species, the Saccharomycotina species Saccharomyces cerevisiae and C. albicans as well as the Taphrinomycotina species Schizosaccharomyces pombe (Kosman, 2003).

The A. fumigatus genome encodes 15 putative MR (Nierman et al., 2005). Combining phylogenetic analysis, genome-wide expression profiling, gene deletion analysis with subsequent phenotyping and biochemical analysis demonstrated that the MR encoded by AFUA_1G17270, termed FreB, is involved in adaptation to iron starvation and functions most likely in RIA.

Materials and methods

Fungal strains and growth conditions

The fungal strains used are listed in Table 1. All strains used in this study were grown at 37 °C in Aspergillus minimal medium (AMM) according to Pontecorvo et al. (1953) containing 1% glucose as carbon source and 20 mM glutamine as nitrogen source. For iron-replete (+Fe) conditions, FeCl3 was added to a final concentration of 30 μM, copper-replete (+Cu) conditions contained 16 μM CuSO4. For iron (−Fe) and copper (−Cu) starvation, addition of iron and copper, respectively was omitted. Liquid cultures were conducted in 0.5 l Erlenmeyer flasks inoculated with 108 conidia and shaken with 200 rpm at 37 °C. The iron chelator bathophenantroline disulfonate (BPS) was used in a final concentration 0.2 mM for plate assays.

Table 1

A. fumigatus strains used in this study.

Strain	Genotype	Reference
Wild type (wt)	ATCC46645	American type culture collection
ΔsidA	ATCC46645; sidA::hph	Schrettl et al. (2004)
ΔsreA	ATCC46645; sreA::hph	Schrettl et al. (2008)
ΔfreB	ATCC46645; freB::ptrA	This work
ΔfreBΔsidA	ΔsidA; freB::ptrA	This work
freBCΔsidA	ΔsidA; ΔfreB, freB, ble	This work

Manipulation of nucleic acids and strain construction

In order to delete the freB gene in ATCC46645 (wt) and ΔsidA strains (Schrettl et al., 2004), the bipartite marker technique was used (Nielsen et al., 2006). Therefore, fungal strains were co-transformed with two DNA fragments, each containing overlapping, but incomplete fragments of the pyrithiamine resistance-conferring ptrA gene fused to the freB 5′- and 3′-flanking sequences. The freB 5′-flanking region (1.1 kb) was PCR-amplified from genomic DNA using primers oFreb1 and oFreB3r, with an added HindIII restriction site. For the amplification of the 3′-flanking region (0.9 kb) primers oFreB4 and oFreB6r were employed. Following gel-purification, the 5′- and 3′-flanking regions were digested with HindIII and PstI, respectively. The ptrA cassette was released from plasmid pSK275 by digestion with HindIII and PstI (2.0 kb) and subsequently ligated to the 5′- and 3′-flanking regions, respectively. The transformation constructs A and B containing fusions of overlapping ptrA cassette fragments with the freB 5′- and 3′-flanking regions, respectively, were PCR amplified from the ligation products using primer pairs oFreB2/oPtrA1 (2.2 kb) and oFreB5r/ptrA2 (2.3 kb). For transformation of fungal strains, both constructs A and B were simultaneously used (Kubodera et al., 2000). Due to the use of nested primers and restriction enzyme digestions, the final length of the freB 5′- and 3′-flanking regions was 1.0 kb and 0.8 kb, respectively. This strategy deleted the region +230 to +2356 according to the translation start site (Fig. S1).

For reconstitution of freB in ΔfreBΔsidA, a 5.6 kb PCR-fragment generated with primers oFreB7 and oFreB5r was used in co-transformation with the plasmid pAN8.1 carrying the phleomycine resistance cassette (Punt and van den Hondel, 1992).

Transformation was carried out as described previously (Schrettl et al., 2007). For selection of transformants, 0.1 μg ml−1 pyrithiamine (Sigma) or 0.04 μg ml−1 phleomycine were used. Screening of transformants was performed by PCR and confirmed with Southern blot analysis. The hybridization probes for Southern blot analysis of ΔfreB and freB strains were generated by PCR using the primers oFreB 1 and oFreB3r and oPtrA1 and otrA2, respectively (Fig. S1 and Table S1).

Northern blot analysis and nucleic acid manipulations

RNA was isolated using TRI reagent (Sigma). For Northern analysis, 10 μg of total RNA were analyzed as described previously. Hybridization probes and primers are listed in Table S1. For extraction of genomic DNA, mycelia were homogenized and DNA was isolated according to Sambrook et al. (1992).

Analysis of siderophores

Analysis of siderophores was carried out by reversed phase HPLC as described previously (Oberegger et al., 2001). Moreover, to quantify extracellular or intracellular siderophores, culture supernatants or cellular extracts were saturated with FeSO4 and siderophores were extracted with 0.2 volumes of phenol. The phenol phase was separated and subsequent to addition of five volumes of diethylether and 1 volume of water, the siderophore concentration of the aqueous phase was measured photometrically using a molar extinction factor of 2996/440 nm (M−1 cm−1).

Ferrireductase assay

Ferrireductase activity was measured as described previously by Nyhus et al. (1997). The assay relies on the formation of a red BPS–Fe2+ (ε535 nm = 22.14 M−1 cm−1) from colorless HEDTA–Fe3+. The assay contained in a total volume of 2.0 ml, 1.0 ml of fungal culture including mycelia and culture, 1.0 mM BPS, and 1.0 mM Fe3+HEDTA. Formation of BPS–Fe2+ was measured after incubation for 1 h at 37 °C in the dark and mycelia were separated by centrifugation prior to reading the absorbance. BPS–Fe2+ produced in the same time frame by the culture supernatant (e.g. due to production of extracellular reductants) was subtracted.

Phylogenetic analysis

Putative MR were identified in fungal genomes by TBLASTN and psiBLAST searches (Altschul et al., 1997). The identified protein sequences were used for a clustalW alignment with the Pasteur bioweb.2 database (http://mobyle.pasteur.fr). Based on this alignment a rooted tree was generated using the Pasteur bioweb.2 database (Neron et al., 2009).

Oxidative stress sensitivity assay

Susceptibility to reactive oxygen species was determined as described previously (Sugareva et al., 2006). 107 A. fumigatus conidia were mixed with 10 ml −Fe-AMM top agar and poured onto −Fe-AMM plates. In the middle of the plate a hole of 5 mm in diameter was pricked out and filled with 100 μl of a 100 mM H2O2 solution. The diameter of the growth inhibition zone was measured after incubation for 24 h at 37 °C.

Measurement of intracellular reactive oxygen species

100 μl −Fe-AMM and +Fe-AMM, respectively, were inoculated with 3 × 103 conidia in 96 well plates. After 18 h of incubation at 37 °C, H2O2 was added to a final concentration of 2.0 mM; in control experiments, H2O2 treatment was omitted. After incubation for 30 min, 6-carboxy-2′,7′-dichlorofluorescin diacetate (DCF-DH) was added and fluorescence was recorded after incubation for another 30 min using a fluorescence microplate reader (Bertold Mithras LB940) with excitation at 485 nm and emission at 535 nm (Wang and Joseph, 1999). The fluorescence was measured for hydrogenperoxide treated and non-treated samples and normalized to that of the non-treated wt.

Statistical analysis

Results were presented as mean ± standard deviation and statistical significance was defined by t-test with p-values < 0.05 and <0.01, respectively.

Results and discussion

A. fumigatus encodes 15 putative MR

MR contain a ferrireductase-like transmembrane motif (CDD:174502; 273-390) and a NOX-Duox-like FAD/NADP-motif (CDD:99783; 451.741). BLASTP and TBLASTN homology searches identified 15 putative MR-encoding genes in the A. fumigatus genome and 10 in the genome of Aspergillus nidulans (Galagan et al., 2005; Nierman et al., 2005), a close A. fumigatus-relative, which however lacks homologs to FtrA and FetC and consequently RIA (Eisendle et al., 2003; Haas et al., 2008). The numbers of putative MR encoded by the genomes of different fungal species are quite different, e.g., nine in S. cerevisiae,. sixteen in C. albicans, two in S. pombe, six in U. maydis, and six in C. neoformans. MR functions have been analyzed in most detail in S. cerevisiae (Philpott and Protchenko, 2008; Rees and Thiele, 2007; Singh et al., 2007); its nine putative MR are termed Fre1-8 and YGL160W. Fre1-4 are involved in RIA. Fre1 and Fre2 comprise the majority of surface reductase activity; they are required for growth on media that contain low concentrations of Fe3+-salts, and can catalyze the reductive release of iron from a variety of siderophores. Because Fe2+ has a low affinity for siderophore ligands, reduction of siderophore-bound Fe3+ results in the release of Fe2+, which can be taken up by the Fe2+-specific transporters. Fre3 and Fre4 can catalyze the reductive uptake of iron bound to hydroxamate siderophores and rhodotorulic acid, respectively. Fre1p and Fre2p are also copper reductases, converting Cu2+ to usable Cu1+ (Georgatsou et al., 1997). Fre6 localizes to the vacuolar membrane, where it functions in the reductive transport of iron and copper from the vacuole to the cytosol. The functions of the other S. cerevisiae MR are unknown. Expression of the genes encoding Fre1-6 is induced during iron starvation. The gene encoding Fre1 is additionally and that encoding Fre7 exclusively induced by copper starvation. S. pombe possesses two putative MR, of which one, termed Frp1, is transcriptionally repressed by iron and essential for RIA (Roman et al., 1993). The C. albicans genes encoding the MR Cfl1 and Cfl95 are repressed by iron and can restore reductase activity to the orthologous S. cerevisiae mutant (Hammacott et al., 2000; Knight et al., 2002, 2005). Furthermore, the transcript levels of the C. albicans genes encoding the MR Frp1 and Fre2 are upregulated by alkaline growth conditions and Frp1 is induced additionally by iron starvation (Baek et al., 2008). Additionally, iron repression has been reported for the genes encoding the following fungal MR: A. nidulans FreA, C. neoformans Fre1, U. maydis Fer9.

Phylogenetic analyses of the MR of these five ascomycetous and two basidiomycetous species revealed that the iron-repressed MR of S. cerevisiae (Fre1-6) and 11 of the 16 MR of C. albicans are in-paralogs, i.e. they evolved in Saccharomycotina after the split from the other subpyhla (Fig. 1). Moreover, the MR from different species with proven function in RIA (S. cerevisiae Fre1-5 and S. pombe Frp1) or showing upregulation during iron starvation (C. neoformans Fre1 and A. nidulans FreA) are only distantly related. These data indicate that prediction of a function in RIA solely based on evolutionary relation is impossible.

Fig. 1

Phylogentic analysis (rooted neighbor joining tree) of MR from 5 asomycetous and two basidiomycetous species. For MR previously analyzed functionally or transcriptionally (see Section 3.1) the protein names are given. Abbreviations are the following: AFUA. A. fumigatus; AN. A. nidulans; Ca. C. albicans; CNAG. C. neoformans; Sc. S. cerevisiae; SPBC. S. pombe; UM. U. maydis.

Expression of freB is repressed by iron in a SreA-mediated way

Genome wide transcriptional profiling indicated SreA-dependent down-regulation of the putative A. fumigatus MR-encoding gene AFUA_1G17270, termed freB, in a shift from iron depleted to iron-replete conditions (Schrettl et al., 2008). SreA is a transcriptional repressor of genes involved in both siderophore-mediated iron uptake and RIA during iron-replete conditions. Northern blot analysis confirmed SreA-dependent repression by iron of freB similar to the known SreA target mirB (Fig. 2A), which encodes a siderophore transporter (Haas et al., 2008). In contrast to many other SreA target genes, however, freB is not genomically clustered with other iron-regulated genes (Schrettl et al., 2008). In contrast to iron starvation, copper starvation did not upregulate the freB transcript level (Fig. 2A), which makes a role in copper metabolism rather unlikely. Together with several known genes involved in high-affinity iron uptake, freB was previously found to be upregulated during initiation of infection in a murine model of invasive pulmonary aspergillosis, which stresses its role in iron metabolism in vivo (McDonagh et al., 2008).

Fig. 2

Expression of freB is repressed by iron via SreA but not by copper (A) and copper starvation transcriptionally upregulates high-affinity iron uptake in ΔfreBΔsidA (B). For Northern analysis A. fumigatus strains were grown for 24 h in liquid flask cultures under iron/copper-replete (++), iron depleted (−Fe), and copper depleted (−Cu) conditions. The SreA target genes sidA, mirB, and fetC encode the siderophore-biosynthetic ornithine monooxygenase, a siderophore transporter, and the RIA component ferroxidase, respectively (Haas et al., 2008). Ethidium bromide-stained rRNA is shown as control for loading and quality of RNA.

FreB is involved in adaptation to iron starvation and the siderophore system increases copper starvation resistance

To functionally characterize the freB gene product, the FreB-coding region was replaced by the pyrithiamine resistance marker gene ptrA in A. fumigatus wild type strain ATCC46645 (wt) and the descending ΔsidA mutant strain, which lacks siderophore biosynthesis, as described in 2.1, Table 1 and Fig. S1. The resulting gene deletion strains were termed ΔfreB and ΔfreBΔsidA, respectively. To ensure characterization of gene deletion-specific effects, a single functional copy of freB was homologously integrated at the freB locus in ΔsidAΔfreB, yielding freBΔsidA. In all tests performed, freBΔsidA behaved as ΔsidA (Table 2 and data not shown). Deletion of freB in ΔsidA was carried out to block possible compensation of FreB-deficiency by the siderophore system, as we found previously that inactivation of RIA does not lead to growth defects unless carried out in a ΔsidA background (Schrettl et al., 2004).

Table 2

FreB-deficiency decreases liquid growth rate during iron starvation (−Fe) but not copper starvation (−Cu). Biomass production was normalized to that of the wt under the respective growth conditions. The biomass production of the wt was 1150.4 ± 27.6, 873.2 ± 14.8, and 499.1 ± 13 mg in +Fe + Cu (30 μM Fe and 16 μM Cu), −Cu, and −Fe, respectively. The given values are the mean ± STD of four biological replicates. Statistical significance was analyzed by t-test (*p < 0.05; **p < 0.01) comparing ΔsidA and ΔfreB with wt but ΔfreBΔsidA and freBΔsidA with ΔsidA.

	−Fe	−Cu	+Fe + Cu
ΔfreB	92.6 ± 2.4**	98.0 ± 1.7**	101.7 ± 2.9
ΔsidA	80.9 ± 2.7**	75.6 ± 1.6**	90.7 ± 3.4*
ΔfreBΔsidA	37.0 ± 1.7**	70.0 ± 0.9**	86.0 ± 2.3**
freBcΔsidA	73.7 ± 2.5	73.9 ± 0.7	91.6 ± 2.1

Based on the similarity of FreB with MR, we compared the biomass production of the wt and mutant strains in response to iron and copper availability in submersed liquid cultures (Table 2). Biomass production of ΔfreB resembled that of the wt during iron and copper sufficiency (+Fe + Cu) and copper starvation (−Cu) but was 7.4% decreased during iron starvation (−Fe). Deficiency in siderophore-mediated iron uptake by SidA-deficiency impaired the biomass production to a higher degree during copper starvation (−24.4%) than during iron starvation (−19.1%). The increased sensitivity of ΔsidA to copper starvation is most likely explained by the fact that iron supply of this mutant relies solely on RIA, which is copper dependent. In contrast, the wt can alternatively acquire iron via the copper-independent siderophore system. Compared to ΔsidA, FreB-deficiency in ΔsidA had only a mild effect during iron and copper sufficiency (−4.7%) as well as copper starvation (−5.6%) but biomass production was reduced by 43.9% during iron starvation. These data indicate that FreB is crucial mainly for adaptation to iron starvation but plays no significant role during copper starvation, which is consistent with the freB expression pattern (see Fig. 2A). The subtle phenotype of ΔfreB compared to ΔfreBΔsidA indicates that FreB-deficiency can be compensated by the siderophore system. The latter is underlined by a 23% increase of TAFC production by ΔfreB compared to wt after 24 h of submersed growth during iron starvation (data not shown). Previously, compensatory upregulation of TAFC production was observed to be caused also by a lack of the RIA component FtrA (Schrettl et al., 2004), which implicates FreB function in RIA as well.

Assaying radial growth rate with classical plate tests (Fig. 3A), i.e. starting from point-inoculated 104 conidia, SidA-deficiency was found to drastically decrease radial growth during starvation for either iron or copper, which again demonstrates that the siderophore system confers resistance against starvation for iron and copper as seen in liquid cultures (see Table 2). FreB-deficiency did impact the growth pattern in neither wt nor ΔsidA backgrounds (Fig. 3A). In plate growth assays starting from single conidia (Fig. 3B), SidA-deficiency was also found to drastically decrease the colony size unless supplemented with high iron concentrations (hFe, 1.5 mM). FreB-deficiency did not influence radial growth in the wt background but blocked colony formation in the ΔsidA background during iron starvation (−Fe).

Fig. 3

SidA-deficiency results in sensitivity to starvation for copper and iron, respectively, (A) and FreB-deficiency results in the inability of colony formation from single conidia during iron starvation (B). (A) 104 conidia of the respective strain were point-inoculated on AMM plates containing the different iron or copper concentrations (+Fe, 30 μM; hFe, 1.5 mM; + Cu, 16 μM) and were incubated for 48 h at 37 °C. (B) About 20 conidia were plated on copper-replete media containing the indicated iron concentration. BPS inhibits strains lacking the siderophore system by blocking of RIA (Schrettl et al., 2004). Copper starvation blocks formation of the green conidial pigmentation resulting in yellow pigmentation.

To further analyze the link between iron and copper metabolism, expression of components of the two high-affinity iron uptake systems was analyzed at the transcriptional level in wt, ΔfreB, ΔsidA, and ΔfreBΔsidA (Fig. 2B). Iron starvation, but not copper starvation, upregulated the transcript levels of sidA, mirB and the ferroxidase-encoding fetC, which is involved in RIA (Schrettl et al., 2004), in wt and ΔfreB. ΔsidA displayed a similar expression pattern but lacked sidA transcripts due to the deletion of this gene. In contrast to ΔsidA, mirB and fetC transcript levels were upregulated in ΔfreBΔsidA not only by iron starvation but also copper starvation. The most likely explanation is that the copper requirement of this mutant is increased and affects iron acquisition due to impaired siderophore-mediated iron uptake and hampered RIA due to FreB-deficiency.

Taken together, these data underline a crucial role of FreB in adaptation to iron starvation and reveal that the siderophore system increases resistance against copper starvation. Moreover, the different growth assays displayed varying sensitivity in detection of defects in adaptation to iron starvation.

FreB-deficiency reduces ferrireductase activity

Iron-replete liquid cultures of the wt lacked measurable surface ferrireductase activity (data not shown). In contrast, iron starvation induced ferrireductase activity in the wt indicating iron-regulation at transcriptional or enzymatic level (Fig. 4A). ΔsidA displayed a 3.6-fold higher ferrireductase activity probably to compensate the loss of siderophore-mediated iron uptake. FreB-deficiency decreased the ferrireductase activity by 33% in wt and 73% ΔsidA backgrounds. These data demonstrate that FreB accounts for a significant part of the ferrireductase activity in wt and the major ferrireductase activity in the absence of the siderophore system.

Fig. 4

FreB-deficiency reduces ferrireductase activity (A), increases sensitivity to hydrogen peroxide (B), and increases cellular reactive oxygen species in particular in the absence of siderophore biosynthesis (ΔfreBΔsidA). Ferrireductase activity was measured after 24 h of submersed growth at 37 °C during iron starvation and normalized to that of the wt. Hydrogen peroxide sensitivity was analyzed after 48 h of growth in plate diffusion assays. The level of cellular reactive oxygen species was quantified without (open bars) and with (grey bars) hydrogen peroxide treatment by the dichlorofluorescein assay, in which fluorescence is proportional to the level of reactive oxygen species. The given values are the mean ± STD of three biological replicates. Statistical significance was analyzed as described in Table 2 (*p < 0.05; **p < 0.01).

FreB-deficiency increases sensitivity to hydrogen peroxide

Plate diffusion assays revealed that deficiency in either FreB or SidA slightly increases sensitivity to hydrogen peroxide. Simultaneous deficiency in both FreB and SidA resulted in substantial increased sensitivity to hydrogen peroxide (Fig. 4B). These data underline that the two high-affinity iron acquisition systems, siderophore-mediated iron uptake and RIA, are mutually able to partially compensate the individual inactivation.

To alternatively quantify the cellular oxidative stress, the dichlorofluorescein assay (Wang and Joseph, 1999), which measures the oxidation of non-fluorescent dichlorofluorescin derivates into fluorescent dichlorofluorescein derivates by reactive oxygen species, was applied. In this assay, the emitted fluorescence is directly proportional to the cellular level of reactive oxygen species. Compared to ΔsidA, the fluorescence of the ΔfreBΔsidA was increased 1.5-fold and after hydrogen peroxide treatment even 1.8-fold (Fig. 4C). Compared to wt, the ΔfreB and ΔsidA strains displayed only a mild, statistically insignificant, increase in fluorescence. These data are in good agreement with the hydrogen peroxide resistance of the strains.

The increase in both oxidative stress and sensitive to hydrogen peroxide of ΔfreBΔsidA is most likely due to the decreased iron supply, which deranges cellular metabolism and in particular impairs iron-dependent enzymes involved in detoxification of oxidative stress, e.g. catalases and peroxidases, which require heme-iron for activity.

Conclusions

This study identified the first ferrireductase involved in RIA of a filamentous fungus. Moreover, this work demonstrated that copper-independent siderophore-mediated iron uptake increases resistance of A. fumigatus against copper shortage. In contrast, iron acquisition of siderophore-lacking fungal species such as S. cerevisiae, C. albicans and C. neoformans relies mainly on copper-dependent RIA. Inactivation of the siderophore system and, in particular, of both the siderophor system and FreB increased sensitivity to copper shortage also in A. fumigatus underlining the interconnection of these two metals. Previously, iron starvation was also found to impact zinc metabolism (Yasmin et al., 2009). Iron starvation downregulates various iron-dependent pathways to spare iron to prolong survival (Hortschansky et al., 2007; Schrettl et al., 2010). MR like FreB are heme-iron-dependent enzymes and therefore their upregulation during iron starvation demonstrates metabolic prioritization of available iron during this condition.