Fungal siderophore metabolism with a focus on Aspergillus fumigatus: impact on biotic interactions and potential translational applications

Fungi colonize an immense range of habitats involving mutualistic, competitive, and pathogenic relationships with other organisms [1–3]. Due to their diverse survival mechanisms and unique attributes, fungi thus play important roles in recycling of organic material, in food production, in biotechnology as producers of valuable primary and secondary metabolites, as well as in medicine and agriculture as producers of toxins and as pathogens for animals and plants. A central prerequisite for habitat colonization is uptake of nutrients including iron which is an essential trace element for all eukaryotes and nearly all prokaryotes. This metal is one of the most abundant elements on earth. However, its bioavailability is very low in most biological niches as it is easily oxidized by atmospheric oxygen into hardly soluble ferric complexes such as ferric hydroxides [4]. Furthermore, fungal pathogens of animals and plants are confronted with iron limitation during host infection [5]. On the other hand, excessive iron uptake is toxic as free iron causes oxidative stress [6,7]. Therefore, fungi have evolved different, highly controlled, high-affinity iron acquisition strategies including utilization of heme, reductive iron assimilation (RIA), and siderophore-mediated iron acquisition (SIA) [5,8]. For RIA, ferric iron is reduced by membrane-localized metalloreductases such as FreB and reoxidized and taken up by the FetC/FtrA protein complex in Aspergillus fumigatus [5]. Siderophores are low molecular mass (about 1 kD), ferric iron (Fe³⁺)-specific chelators employed by bacteria and fungi for iron acquisition and by fungi also for intracellular iron handling [5,9]. The presesnt review focuses on SIA and mainly on A. fumigatus, as SIA has been studied in most detail in this mold. A. fumigatus has been shown to employ RIA and SIA but appears to lack efficient use of heme as iron source [10]. Control of iron uptake and storage appears to be the major mechanism to maintain fungal iron homeostasis as no excretory mechanism for iron has been identified so far. Iron detoxification relies mainly on vacuolar iron deposition as shown in Saccharomyces cerevisiae and A. fumigatus; the transporter is termed CccA in A. fumigatus, whereby it is unknown if vacuolar stored iron can be reused in A. fumigatus [11]. The strategies for iron uptake and storage of A. fumigatus are summarized in Figure 1.

Most Ascomycota and Basidiomycota species produce hydroxamate-type siderophores displaying a remarkable species-specific structural variety [5]. However, there are prominent exceptions lacking siderophore biosynthesis such as the entire Saccharomycotina clade, including Saccharomyces cerevisiae and Candida albicans, as well as the basidiomycete genus Cryptococcus spp. [5]. Mucoromycota lack hydroxmate-type siderophores but produce a carboxylate-type siderophore, termed rhizoferrin, originally isolated from Rhizopus microsporus [12,13]. Rhizoferrin displays a significantly lower affinity to iron compared with hydroxamates [12]. Rhizoferrin biosynthesis has been shown to depend on a nonribosomal peptide synthetases (NRPS)-independent synthetase (NIS) in Rhizopus delemar [14]. Apart from that, little is known about rhizoferrin-mediated iron uptake. Fungal hydroxamate-type siderophores are grouped into four structural distinct types: fusarinines, coprogens, ferrichromes, and rhodotorulic acid; representatives of each family are shown in Figure 2A. Notably, the terms ferrichrome and coprogen refer to specific siderophores in their structural type. Unfortunately, the siderophore name refers in some cases to the iron-complexed form (e.g., ferrichrome) and in some cases to the metal-free form (e.g., coprogen). Siderophores are synthesized in the metal-free form but taken up only as metal complex. Detailed chemistry of the structural diversity of fungal siderophores has been reviewed previously [9,15]. In addition to hydroxamate- and carboxylate-type siderophores, bacteria produce also catecholate-, carboxylate-, and mixed-type siderophores [9,16]. Plants employ rather simple iron chelators such as nicotianamine, mugineic acid family phytosiderophores, and citrate for iron mobilization [17].

Siderophore biosynthesis has been studied in most detail in A. fumigatus (Figure 2B). This mold secretes two fusarinine-type siderophores, fusarinine C (FsC), and triacetylfusarinine C (TAFC) to capture environmental iron [18,19]. Moreover, it employs two ferrichrome-type siderophores, hyphal ferricrocin (FC), and conidial hydroxyferricrocin (HFC) for intracellular handling of iron such as transport of iron through conidiophores for conidiation and conidial iron storage [18,20,21]. A recent study indicated that FC is also secreted and plays a particular role for iron uptake during germination [22]. The common initial step for biosynthesis of all hydroxamate-type siderophores is formation of N⁵-hydroxyornithine from ornithine, which is catalyzed by the monooxygenase SidA [10]. Subsequently, the pathways for biosynthesis of different hydroxamate types split due to incorporation of different acyl-groups, which generates the iron-chelating hydroxamate group. For biosynthesis of fusarinine- and coprogen-type siderophores, the transacylase SidF transfers anhydromevalonyl-CoA to N⁵-hydroxyornithine [18,23,24]. Anhydromevalonyl-CoA is derived from mevalonate by the mevalonyl-CoA ligase SidI and the mevalonyl-CoA hydratase SidH [18,23,24]. SidI, SidH, and SidF are localized in peroxisomes, while the other siderophore biosynthetic enzymes are assumed to function in the cytosol [25]. The linkage of three N⁵-anhydromevalonyl-N⁵-hydroxyornithine to cyclic FsC or linear coprogens is mediated by NRPSs displaying a similar architecture, e.g., A. fumigatus SidD for FsC and Cochliobolus heterostrophus Nps6 for coprogens [18,26,27]. FsC is then triacetylated by SidG to form more stable TAFC [18]. Derivatization of coprogen B, including acetylation and methylation, allows for the diversity of coprogen-type siderophores. For synthesis of ferrichrome-type siderophores, which are cyclic hexapeptides, N⁵-hydroxyornithine is acylated, e.g., in A. fumigatus acetylated to N⁵-acetyl-N⁵-hydroxyornithine by the transacetylase SidL and another yet unknown enzyme [28]. Subsequently, the hexapeptide FC is assembled by the NRPS SidC from three N⁵-acetyl-N⁵-hydroxyornithine, two glycine, and one serine residue [5,24]. Conidial HFC is formed from FC by a single hydroxylation by an as yet unknown enzyme [18]. Figure 3 displays some ferrichrome-type siderophores containing different amino acids and acyl-groups. Notably, Epichloënin A produced by Epichloë festuce is an atypical ferrichrome-type siderophore consisting of eight amino acid residues [29]. The presence of three hydroxamate groups, as found in most hydroxamate-type siderophores, allows formation of hexadentate structures, which provide the highest affinity for ferric iron [9]. The mechanism of cellular siderophore export has not been characterized yet.

Uptake of hydroxamate-type siderophore-iron chelates in fungi is mediated by members of the siderophore iron transporter (SIT) family, which is a subfamily of the major facilitator superfamily. SITs are exclusively found in the fungal kingdom. Most species from all fungal phyla (Ascomycota, Basidiomycota, Mucoromycota, and Chytridiomycota) possess SITs [5,30]. Remarkably, species that lack siderophore production have preserved the ability to utilize siderophores. The ability to utilize xenosiderophores (siderophore-types other than self-produced ones) saves energy and most likely plays a role in microbial interaction (see below).

Functional characterization of SITs started in S. cerevisiae, which possesses four SITs with different substrate specificity: Sit1p/Arn3p for ferrioxamine B (a bacterial hydroxamate-type siderophore), Arn1p for ferrichromes, Taf1p/Arn2p for TAFC, Enb1p/Arn4p for the bacterial catecholate-type siderophore enterobactin [31]. In contrast, C. albicans contains only a single broad-range substrate-specific SIT, CaArn1p/CaSit1p [31]. Phylogenetic analysis revealed that all Saccharomycotina SITs are more similar to each other than they are to SITs from other fungal species [19,30], which indicates that these transporters arose after the divergence from the other species. Consequently, the substrate specificity of non-Saccharomycotina SITs, such as those of mold species, cannot be predicted on the basis of sequence similarity to S. cerevisiae SITs.

A. fumigatus can utilize a variety of hydroxamate-type siderophores, including several ferrichromes, endogenously secreted FsC and TAFC, several ferrioxamines, and coprogens, although the latter are utilized only poorly [19]. In contrast, A. fumigatus was found to be unable to utilize the hydroxamate-type siderophores basidiochrome and rhodotorulic acid, the bacterial catecholate-type siderophore enterobactin, rhizoferrin, and the mixed-type siderophores ornibactin and schizokinen [19]. Recent studies revealed the substrate specificities of four SITs (Sit1, Sit2, MirB, MirD) of A. fumigatus [19,32,33]. Sit1 and Sit2 were found to be essential for the utilization of coprogen- and ferrichrome-type siderophores, displaying both redundancy and exclusivity depending on the amino acid residues and acyl-groups present in the respective hexapeptide [19], which is summarized in Figure 3. In detail, (i) both Sit1 and Sit2 accept serine and glycine in positions R₁ and R₂ and acetyl as acyl-group in R₄–R₆; (ii) Sit2 but not Sit1 accepts anhydromevalonyl as acyl-group in positions R₄–R₆; (iii) Sit2 does not distinguish between cis- and trans-anhydromevalonyl as acyl-group in positions R₄–R₆; (iv) Sit1, and to a lesser extent Sit2, accept methylglutaconyl as acyl-group in positions R₄–R₆; (v) methylglutaconyl as acyl-group in R₄–R₆ significantly decreases uptake efficacy in comparison with anhydromevalonyl; and (vi) Sit1 accepts asparagine, leucine, and D-phenylalanine in positions R₁, R₂, and R₃, while at least one of these amino acid residues disturbs recognition by Sit2 and therefore uptake of the ferrichrome-type antifungal VL-2397 depends exclusively on Sit1 [19,34]. These results demonstrate that both the amino acid residues in positions R₁–R₃ as well as the acyl-groups in positions R₄–R₆ impact recognition of ferrichrome-type siderophores. Similar to ferrichrome A, the utilization efficacy of coprogen-type siderophores was low despite the fact that these siderophores were accepted by both Sit1 and Sit2 [19]. Furthermore, Sit1 was shown to be the exclusive transporter of bacterial ferrioxamines [19], and that the transport efficacy of linear ferrioxamines is impacted by their charge and consequently the environmental pH [35]. Acquisition of TAFC was found to depend exclusively on MirB and that of FsC mainly on MirD [33]. Taken together, for every siderophore known to be utilized by A. fumigatus, a corresponding major transporter has been identified. Sit1, Sit2, and MirD were found to be dispensable for virulence in murine models of pulmonary aspergillosis [19,32,33]. In line with TAFC being the major secreted siderophore for iron acquisition, MirB is important for virulence [33]. The loss of MirB not only impaired uptake of TAFC but additionally caused an autoinhibitory effect by decreasing the bioavailability of environmental iron due to chelation by a now futile siderophore. Moreover, the Aspergillus nidulans SIT MirA was indicated to transport bacterial catecholate-type siderophore enterobactin [36]. The insights in the substrate specificity of SITs will help to reveal the molecular basis of substrate recognition.

Phylogenetic analysis of 38 SITs from 13 fungal species shown in Figure 4 demonstrated that the four functionally characterized A. fumigatus SITs belong to different subclades [19]. As mentioned above, all SIT family members of the Saccharomycotina species S. cerevisiae, C. albicans, and C. glabrata are closely related building a sister clade with the ‘Sit1 clade,’ which indicates a common origin. Within the A. fumigatus Sit1 subclade, Fusarium graminearum Sit1 has been shown to transport ferrichrome and ferrioxamine B and Cryptococcus neoformans Sit1 was found to transport ferrioxamine B [37,38], which underlines a link between phylogenetic clustering substrate specificity. Despite overlapping substrate specificities, Sit1 and Sit2 are only distantly related, while MirB and MirD are localized in sister clades, indicating coevolution. The latter finding might be related to the fact that TAFC is biochemically derived from FsC, requiring only a single enzymatic step, i.e., triacetylation catalyzed by SidG [18]. Among the species analyzed, MirB was found to be conserved in Fusarium oxysporum, Aspergillus lentulus and A. nidulans. Importantly, the A. nidulans MirB homolog was shown to transport TAFC when heterologously expressed in S. cerevisiae [36]. Another subclade in the phylogenetic analysis comprises homologs of a fifth putative siderophore transporter of A. fumigatus, MirC, which might play a function in FC biosynthesis [39]. Notably, two SITs shown in the phylogenetic analysis have been shown to accept nonsiderophore substrates; as S. cerevisiae Gex2 and Schizosaccharomyces pombe Str3 have been reported to transport glutathione and heme, respectively [40,41]. In contrast, S. pombe Str1 was found to transport ferrichrome [42]. Taken together, these data indicate the value of phylogenetic analysis for prediction of SIT substrate specificities outside of the Saccharomycotina subclade.

Due to the high affinity of siderophores to iron, intracellular mechanisms to release the chelated iron from siderophores are required. Three esterases have been identified, which intracellularly hydrolyze siderophores with narrow substrate specificity: A. fumigatus EstB for TAFC, A. fumigatus SidJ for FsC, and A. nidulans EstA for enterobactin [43–45]. In bacteria, iron is liberated from the ferric siderophore complex through reduction to ferrous iron by enzymes termed ferric reductases [46]. Such a mechanism has not been identified in the fungal kingdom yet.

Similar to other high-affinity iron acquisition mechanisms, SIA is induced by iron limitation and repressed by iron to avoid excessive iron uptake. A. fumigatus employs two iron-sensing transcription factors, termed SreA and HapX [5,47]. SreA harbors two Cys₂Cys₂ GATA-type zinc fingers, which recognize the consensus DNA sequence ATCWGATAA, separated by a cysteine-rich region (CRR), which most likely mediates iron sensing. HapX comprises several phylogenetically conserved domains: (i) a bZIP-type DNA-binding domain, (ii) a Hap4-like domain (Hap4L) for physical interaction with the CCAAT-binding complex (CBC, termed Hap complex in S. cerevisiae), and (iii) four CRR, whereby two CRR (CRR-A and CRR-B) have been implicated in iron sensing. During iron sufficiency, SreA transcriptionally represses high-affinity iron uptake, including RIA and SIA. During iron starvation, HapX represses iron-consuming pathways such as respiration, heme biosynthesis, TCA cycle, and vacuolar iron deposition to spare iron. Furthermore, HapX transcriptionally activates SIA [48]. Remarkably, iron excess converts HapX into an activator of iron-dependent pathways, particularly vacuolar iron deposition. Inactivation of both HapX and SreA is synthetically lethal, underlining the critical role of iron homeostasis in cellular survival [48,49]. Both HapX and SreA are highly conserved in Ascomycota and Basidiomycota species [5,47]. Interestingly, Saccharomycetaceae, including S. cerevisae but not Candida albicans, lost SreA, conserved only the iron detoxification function of HapX leading to Yap5 proteins, and evolved novel iron regulators termed Aft1/2 [5,47]. Iron sensing in A. fumigatus was shown to depend on mitochondrial but not cytosolic iron sulfur cluster biosynthesis [50]. Most likely SreA and HapX sense the cellular iron status via binding of [2Fe-2S] cluster with their CRR additionally involving the [2Fe-2S] cluster chaperon GrxD and glutathione [5,47,50].

Recent studies indicated that iron shortage is additionally sensed via iron-dependent metabolic pathways. Biosynthesis of the branched-chain amino acids leucine, isoleucine, and valine includes enzymes with iron–sulfur clusters as cofactors and consequently iron limitation leads to accumulation of the pathway intermediate α-isopropylmalate, which post-translationally activates the Zn₂Cys₆-type transcription factor LeuB for feedback activation of leucine biosynthesis [51]. Recently, LeuB was shown to be required for full transcriptional activation of SIA in A. fumigatus indicating integration of metabolic signals in control of iron homeostasis [52,53]. Moreover, the sterol regulatory element-binding protein (SREBP) transcription factor SrbA and the Gal4-type zinc finger protein AtrR were found to mediate activation of high-affinity iron acquisition including RIA and SIA [54,55]. Previously, SrbA and AtrR were found to be essential for sterol-feedback regulation and consequently resistance against triazole drugs as well as for hypoxic growth [56,57]. Consequently, SrbA and AtrR link control of iron acquisition and iron consumption as ergosterol biosynthesis and adaptation to oxygen limitation are iron-dependent pathways [23,58].

Iron is an essential trace element that is limiting in most habitats including plant and animal hosts of fungal pathogens. Moreover, mammalian innate immunity further increases restriction of iron availability to fight infections within ‘nutritional immunity,’ leading to anemia of inflammation [59,60]. Consequently, pathogens evolved strategies to ‘steal’ the iron from their hosts. Thus, competition for iron is a critical battleground that determines the outcome of host–pathogen relationships. In agreement, iron overload decreases mammalian resistance against infection including invasive aspergillosis [61,62]. In healthy individuals, plasma iron accessible for cells is bound to the iron-transporting protein transferrin, whereby transferrin is typically only about 30% iron-saturated. However, pathologic iron excess conditions can exceed the binding capacity of transferrin leading to nontransferrin-bound iron (NTBI), which was found to stimulate the in vitro growth of A. fumigatus in serum from hematopoietic stem cell-transplanted patients [63]. The first link between siderophores and virulence was the association between infection with Mucoromycota species in dialysis patients and the use of desferrioxamine against aluminum overload and/or iron excess, a link that was confirmed in animal models in 1989 [64]. Later on, desferrioxamine was shown to serve as xenosiderophore for these species [65]. However, the finding that siderophore biosynthesis is dispensable for virulence of the maize pathogen Ustilago maydis in 1993 largely diminished the interest in fungal SIA [66]. This view changed significantly, when blocking siderophore biosynthesis by inactivation of SidA was found to render A. fumigatus avirulent in murine models for invasive aspergillosis in 2004, while blocking RIA had no effect [10]. Subsequently, biosynthesis of both fusarinine-type and ferrichrome-type siderophores and uptake of TAFC by MirB were found to be crucial for virulence of this mold [18,33]. Siderophore biosynthesis was also found to be crucial for virulence of several plant pathogenic species including F. graminearum, C. heterostrophus, and Alternaria brassicicola [26,67–69]. These fungal species are necrotrophs, which rapidly kill plant tissue after invasion. In contrast, U. maydis is a biotrophic pathogen that establishes a long-term feeding relationship with its hosts without causing immediate damage. Hemibiotrophs switch after a short initial biotrophic phase to necrotrophy. In the hemibiotrophic plant pathogen Colletotrichum graminicola, SIA was found to be repressed during biotrophic- and activated during necrotrophic growth [70]. In line with the importance of SIA for necrotrophic virulence and RIA for biotrophic virulence, both SIA and RIA were found to be crucial for virulence of C. graminicola but only RIA for virulence of U. maydis [71]. The presence of siderophores modulates the plant immune response and down-regulation of SIA might serve to evade the plant immune response in biotrophic pathogens [72–74]. Recently, the carboxylate-type siderophore rhizoferrin was shown to also play a role in virulence of the Mucoromycota species Mucor lusitanicus [75].

In line with the importance of SIA for virulence of A. fumigatus, metabolic pathways that are required in a nonexclusive manner for SIA were found to be crucial for virulence of A. fumigatus including mitochondrial production of the siderophore precursor ornithine [76], leucine biosynthetic enzymes that control post-translational activation of LeuB and consequently affect iron regulation [53], biosynthesis of riboflavin that is an essential cofactor for the first step of siderophore biosynthesis [77], and biosynthesis of pantothenic acid, which is a precursor of the substrate-binding 4’-phosphopantethine groups of NRPSs including SidD and SidC [77]. Lack of extracellular siderophore biosynthesis or SreA resulted in perturbation of the mutualistic interaction of the endophyte E. festucae and its perennial ryegrass host [29,78], which indicates that siderophores also play a role in symbiotic interactions.

The battle for iron significantly also impacts the interaction of microorganisms leading to beneficial and antagonistic relationships. What is the reason for the stunning structural diversity of siderophores produced by fungi and bacteria? Most likely, the rational is that siderophores produced by a species can either promote or repress the growth of another species in the same habitat depending on its siderophore utilization capacity: growth promotion by improving iron uptake in case of uptake of the siderophore and growth repression by chelation of iron by a siderophore that is not recognized. The optimal competitive strategy would therefore be the production of siderophores that are not recognized by competitors combined with the capacity to utilize as many siderophore structures as possible. This is the most likely explanation for the utilization of xenosiderophores also by siderophore-producing species, e.g., utilization of enterobactin via MirA by A. nidulans [36] or the utilization of ferrichrome- and coprogen-type siderophores, which are produced exclusively by fungal species, by bacteria such as Escherichia coli or Salmonella spp. [79–81]. Similarly, the utilization of siderophores by species that do not produce siderophores such as Saccharomycotina or C. neoformans might be explained by the potential antagonistic activity of siderophores or alternatively to save energy as siderophore-chelated iron represents a soluble energized iron-form. The interplay between A. fumigatus and the bacterium Pseudomonas aeruginosa, which frequently coinfects the airways of cystic fibrosis patients, visualizes the antagonistic activity of siderophores: P. aeruginosa can iron-starve A. fumigatus via iron sequestration by its major siderophore pyoverdine, which cannot be utilized by A. fumigatus [82]. In this competition for iron, siderophore production by A. fumigatus plays an important role in protection against P. aeruginosa [83]. In this respect, it is noteworthy that TAFC was found to inhibit growth of a range of bacterial species, which are obviously not able to utilize this siderophore [84]. The existence of natural siderophore-auxotrophic fungal species demonstrates the beneficial activity of siderophores. Examples are the mucoromycete Pilobolus spp., which requires coprogen or ferrichrome as growth factor [85], the ascomycete Debaryomyces mycophilus, which lives as endosymbiont in the guts of woodlice, and the basidiomycete Tritirachium egenum, which is a mycosymbiont lacking independent high-affinity iron acquisition and depending on xenosiderophore supply during growth in association with Penicillium rugulosum [86,87].

The use of siderophores distinguishes microbial from plant and mammalian cells, which might enable translational applications. The most prominent translational application of siderophores is the bacterial siderophore desferrioxamine B (trade name desferal®), which has been used in clinics for metal chelation therapy in iron or aluminum overload pathologies [88]. However, this therapy led to major problems with patient compliance due to the requirement for long periods of subcutaneous infusions of this orally ineffective drug. Consequently, desferal® was widely replaced by the orally applicable synthetic chelator drug deferasirox [88]. As most siderophores have a high affinity not only for iron but also for some other metals, desferrioxamine B was used as a chelator for radionuclides, initially for Gallium-67 for Single Photon Emission Tomography (SPECT) applications [89]. In recent years, desferrioxamine B has become the chelator of choice to label in particular antibodies with Zirkonium-89 (⁸⁹Zr) for Positron Emission Tomography (PET), so-called immuno-PET imaging [90]. The reported in vivo release of ⁸⁹Zr from the desferrioxamine B chelator has stimulated research to develop more stable chelators for PET imaging applications [91]. Among the developments, the use of FsC as scaffold to attach targeting vectors such as peptides has shown promising results to construct multimeric targeting probes for Gallium-68 (⁶⁸Ga) and ⁸⁹Zr labeling [92–94], as well as ⁶⁸Ga-labeled probes for hybrid imaging [95] and pretargeting [96] applications. Using diacetylfusarinine C (TAFC lacking a single acetyl group allowing conjugation with the amino group) as bifunctional chelator to label an EGFR-targeting ZEGFR:2377 affibody with ⁸⁹Zr revealed advantages over desferrioxamine B [97].

A. fumigatus is the most common human mold pathogen [98]. Invasive infections, occurring mainly in immunosuppressed patients, are rather rare but nevertheless life-threatening [98]. The diagnosis of fungal infections is difficult as it lacks specificity and sensitivity. Recently, TAFC was reported to be an attractive novel biomarker for systemic A. fumigatus infection enabling noninvasive diagnosis in human urine and bronchoalveolar lavage [99,100]. Detection of TAFC in urine is more sensitive than in serum because in murine and rat models, TAFC showed a short half-life in blood due to rapid renal excretion in intact form: within 45 min, about 90% of injected TAFC was found in kidneys and bladder [100]. In an A. fumigatus rat infection model not only TAFC but also FC was detected in urine and FC also in serum [101], which appears to have a longer half-life in serum compared with TAFC [102]. However, sensitive siderophore detection in human and animal samples is based so far on mass spectrometry, which represents a problem in clinics. This might change with optimization of detection with Raman spectroscopy or immunological assays employing a recently developed antibody recognizing TAFC [103,104].

Furthermore, replacing iron in siderophores such as TAFC, ferrioxamine E, and ferrioxamine B by the radionuclide ⁶⁸Ga allowed in vivo imaging of A. fumigatus infection by PET due to the specific uptake and accumulation of the siderophore in fungal cells [35,105,106], providing high specificity for detecting fungal infections [107]. Figure 5 shows PET/computer tomography (CT) images using ⁶⁸Ga–desferrioxamine B for imaging pulmonary aspergillosis in a rat model [35]. Moreover, the conjugation of siderophores and fluorescent dyes enabled the generation of hybrid-imaging compounds, allowing the combination of PET and optical imaging in preclinical aspergillosis models [108]. Remarkably, conjugates of fluorescent dyes and diacetylfusarinine C showed subcellular localization dependent on the fluorescent molecule, i.e., conjugates with nitrobenzoxadiazole (NBD) and Ocean Blue accumulated in vacuoles, conjugates with BODIPY, silicon-rhodamine (SiR), and cyanine 5-carboxylic acid (Cy5) localized to mitochondria, and the conjugate with fluorescein isothiocyanate (FITC) showed cytoplasmic distribution [109]. These results emphasize that interpretation of the cellular fate of siderophores after uptake is problematic based on fluorescent labeling. Nevertheless, the uptake of all these compounds depended on the SIT MirB. These data revealed that SITs tolerate substantial derivatization of their substrate. This is particularly interesting as SITs represent one of few protein families that are unique to the fungal kingdom, i.e., they are not present in prokaryotes or other eukaryotes. Consequently, SITs might allow fungal-specific drug delivery by a Trojan horse approach [110,111], in which toxic compounds are conjugated to siderophores for selective import into fungal cells. TAFC derivatives appear most promising as the importance of MirB in virulence and the autoinhibition caused by its inactivation restrains its mutational inactivation and, consequently, the development of resistance at the level of uptake [33]. Synthesis of such derivatives resulted in compounds with traceable antifungal activities and accumulation in infected as shown by ⁶⁸Ga-labeling and PET [112]. The natural ferrichrome-type antifungal drug VL-2397 (previously termed ASP2397), produced by Acremonium persicinum and containing aluminum instead of iron, was recently shown to require the SIT Sit1 for uptake and activity against A. fumigatus, while its mode of action remains elusive [34]. Notably, there are natural antibacterial siderophore-antibiotic conjugates, termed sideromycins and recently the first synthetic siderophore-antibiotic conjugate, cefiderocol, was FDA-approved to combat multidrug-resistant Gram-negative bacteria [113].

Due to its crucial role in virulence, the siderophore biosynthetic pathway represents a potential target for selective therapeutic intervention. A natural quinone methide, celastrol, was identified as a noncompetitive inhibitor of SidA [114]. Moreover, treatment of fungal keratitis in a murine model by dual topical therapy with the iron chelator deferiprone and statins, which blocks hydroxymethylglutaryl (HMG)-CoA reductase and consequently biosynthesis of isoprenoids and extracellular siderophores (Figure 2B), showed restriction of fungal growth [115].

• Siderophores play a central role in maintenance of iron homeostasis in most fungal species.

• Siderophores display a stunning species-specific structural variety.

• Siderophores play a crucial role mutualistic and antagonistic biotic interactions including fungal virulence in animal and plant hosts.

• Siderophores and siderophore uptake show great potential as biomarkers and for imaging of fungal infections.

• The siderophore system might allow fungal-specific drug delivery by a Trojan horse approach in which toxic compounds are conjugated to siderophores for selective import by fungal cells.

The authors declare that there are no competing interests associated with the manuscript.

This research was funded by the Austrian Science Fund (FWF) [doctoral program Host Response in Opportunistic Infections [HOROS] W1253 (to H.H.) and FWF ZFP30924-B26 (to C.D.)].

⁶⁸Ga

Gallium-68

⁸⁹Zr

Zirkonium-89

CBC

CCAAT-binding complex

CoA

coenzyme A

CRR

cysteine-rich region

CT

computer tomography

FC

ferricrocin

FsC

fusarinine C

NIS

NRPS-independent synthetase

NRPS

nonribosomal peptide synthetase

PET

positron emission tomography

RIA

reductive iron assimilation

SIA

siderophore-mediated iron acquisition

SIT

siderophore iron transporter

TAFC

triacetylfusarinine C