INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Histoplasma capsulatum is the causative agent of histoplasmosis. Although distributed worldwide, H. capsulatum has classically been associated with the U.S. midwest. Infection in areas where the organism is endemic, detected by delayed-type hypersensitivity to H. capsulatum antigens, can occur in 90% of the population. H. capsulatum is thermally dimorphic, existing as a mold in soil and as a yeast in the mammalian host; these morphotypes are also observed in laboratory medium at room temperature and 37°C, respectively. Infection occurs through the inhalation of conidia or mycelial fragments. H. capsulatum then converts to yeast and infects pulmonary macrophages, where it replicates in a phagosomal or phagolysosomal compartment (39). H. capsulatum modulates but does not completely reverse or block phagolysosome acidification and maintains a pH of approximately 6.5 (13, 33, 45). To immunocompromised individuals, histoplasmosis poses a greater risk and results in severe or fatal infection (39). The identification of factors involved in H. capsulatum survival within the host is important for the development of vaccines or chemotherapeutics, as well as for understanding pathogenic mechanisms.

Iron is required for both intracellular and in vitro growth of H. capsulatum. Iron sequestration by gamma interferon-treated macrophages is a specific host defense against H. capsulatum (22); additionally, chelation of Fe(II) or Fe(III) inhibits H. capsulatum growth in macrophages (34, 35). In vitro H. capsulatum growth is enhanced by Fe(II) or Fe(III) salts or human holotransferrin (holo Tf) and is inhibited by Fe(III) chelation (46).

Iron is not readily available in the soil or host. In soil, iron forms chelates or insoluble hydroxides. Siderophores, low-molecular-weight (MW) Fe(III) chelators, are secreted by many bacteria and fungi and are present in soil at physiologically relevant concentrations, e.g., 0.1 to 10 µM for the Streptomyces siderophore ferrioxamine B (FOB) (24). Utilization of foreign ferric siderophores by bacteria and fungi has been documented for both siderophore-producing and siderophore-non-producing microbes (3, 6, 15, 17, 24, 25, 47, 50). During infection most iron is present in host iron-binding compounds such as transferrin (Tf), lactoferrin, hemoglobin, heme, and hemin, but the forms of iron present in the H. capsulatum-containing phagolysosome have not been identified. Tf is a bilobed glycoprotein that can bind two atoms of Fe(III) and can transport Fe(III) into the mammalian cell. When diferric Tf or, less thermodynamically favorably, monoferric Tf binds the Tf receptor, the complex is internalized into an endosome that acidifies. At pH 5.6 one atom of Fe(III) is lost preferentially from the C-terminal iron binding site when diferric Tf is bound to the Tf receptor, while iron is first lost from the N-terminal iron binding site of free diferric Tf. Fe(III) is then exported into the mammalian cytoplasm (2). In human serum Tf is present at 40 µM (27) and is saturated to 20 to 45% (41).

One microbial strategy to circumvent the inaccessibility and toxicity of iron is the secretion of siderophores during iron limitation. Both mold- and yeast-phase H. capsulatum produce hydroxamate siderophores, including coprogens, fusarinines, and dimerum acid (DA); DA is a dihydroxamate derivative of the basidiomycete siderophore rhodotorulic acid (RA), a diketopiperazine of N--acetyl-L-(S)-N--hydroxyornithine (7, 18, 19). H. capsulatum also produces cell surface ferric reducing agent(s), extracellular low-MW nonproteinaceous ferric reductant(s), and extracellular glutathione-dependent ferric reductase (GSH-FR); these agents are induced or derepressed during iron stress but also are produced during growth in iron-replete conditions (46). Here, we present evidence for a relationship between H. capsulatum iron acquisition mechanisms and a potential role for the GSH-FR in iron acquisition in the host and environment.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Fungal strain and growth conditions. H. capsulatum G217B (ATCC 26032) is a clinical isolate of restriction fragment length polymorphism class 2. Stock cultures were maintained in HMM, which has been previously described and is based on the tissue culture medium Ham's F-12, which contains 3 µM FeSO₄ (49). Except where noted, G217B was grown in RPMI 1640 medium (Gibco-BRL, Grand Island, N.Y.) without phenol red or bicarbonate, supplemented with 1.8% dextrose, 25 mM HEPES, and penicillin (10 IU/ml)-streptomycin (10 µg/ml) in broth or gentamicin (15 µg/ml) in agarose, adjusted to pH 7.5 (modified RPMI [mRPMI]). This medium contains no added iron, but the presence of trace contaminating amounts is highly likely. Cultures were grown in mRPMI containing 0.5 µM deferoxamine mesylate (Df; Sigma, St. Louis, Mo.) to obtain a high-MW culture supernatant fraction (46). To detect coproduction of siderophores and ferric reducing agents, an early-log-phase HMM culture was washed with phosphate-buffered saline (PBS) and resuspended in a twofold dilution in mRPMI containing 5 µM Df. After incubation at 37°C for two additional days, iron-depleted cells were washed with PBS and inoculated into RPMI 1640 supplemented with 0.5% dextrose and 25 mM HEPES and treated with 5% (wt/vol) Chelex 100 (Bio-Rad, Hercules, Calif.) for 1 h. Disposable polycarbonate flasks were used with Chelex-treated media; media were unsupplemented or supplemented with 10 µM FeSO₄. Cultures were incubated 7 days, at which point culture turbidity of the iron-replete sample was ca. 1.7-fold that of the iron-limited sample (data not shown). Unless noted, cultures were grown as yeast at 37°C in 5% CO₂.

Growth quantitation assays. Glassware was washed with the alkaline detergent Contrad 70 (Fisher Scientific, Hanover Park, Ill.) and rinsed thoroughly in distilled deionized water prior to use. For growth assays, cells were partially depleted of intracellular iron by overnight incubation in mRPMI containing 10 µM Df as described previously (46). Iron-depleted cells were washed with PBS, and 2 × 10⁶ to 3 × 10⁶ CFU ml¹ (one absorbance unit at 600 nm [A₆₀₀] = 2.24 × 10⁸ CFU/ml [data not shown]) were inoculated into mRPMI. To assay the effect of foreign siderophores, mRPMI was supplemented with 5 µM iron-free or iron-replete siderophore (prepared as described below) or FeCl₃. To assay the utilization of hemin, mRPMI was supplemented with 1 µM hemin [Fe(III)-protoporphyrin IX [PIX]; Frontier Science, Logan, Utah], 1 µM PIX (Frontier Science), or 1 µM FeCl₃. To assay the effect of Tf, mRPMI was supplemented with 1 mM NaHCO₃, a cofactor for Fe(III) binding by Tf (2) and included in all cultures for uniformity and 1 µM 0% (apo), 25%, 50%, or 100% saturated (holo) Tf (Sigma). For host iron sources, the pH of culture supernatant was monitored throughout the growth curve with Emerson colorpHast strips (pH 6.5 to 10.0 and pH 5 to 10; Fisher). Baseline growth in unsupplemented medium was also measured. Culture turbidity was monitored by measuring the A₆₀₀ of 100 µl of culture with a Spectramax 250 microplate spectrophotometer; samples were diluted when necessary to be within the linear range.

Preparation of iron sources. Ferric siderophores were prepared by incubation of iron-free siderophore and FeCl₃ in a 1:1 molar ratio [Fe(III)-siderophore] for Df and ferrichrome (FC) (Sigma) and a 2:3 molar ratio for RA (Frontier Science) at 4°C for 24 h (19). Fe₂DA₃ purified from H. capsulatum G184AR culture supernatant was a generous gift from Dexter Howard (University of California at Los Angeles [UCLA]). For growth assays, partially saturated Tf was prepared from apo Tf and FeCl₃ and incubated in mRPMI supplemented with 1 mM NaHCO₃ for 2 to 3 days at 4°C. For other experiments, partially saturated Tf was prepared by mixing apo and holo Tf and incubating the mixtures at 4°C for 2 to 3 days. Tf saturation was confirmed by Tris-borate-EDTA (TBE)-urea-polyacrylamide gel electrophoresis (PAGE) (see below). Tf saturated primarily at the N-terminal binding site (Fe_N-Tf) was prepared by removing iron from the C-terminal binding site of holo Tf as previously described (5). Hemin and PIX were dissolved in 0.1 M NH₄OH.

Sample preparation. Log-phase yeast cultures were used except as noted. Cells were harvested by centrifugation (1,200 × g, 10 min). Culture supernatants were sterilized through 0.2-µm (pore-size) filters, diafiltered into 50 mM sodium phosphate (pH 7.0), and concentrated 100-fold by ultrafiltration with Ultrafree-15, 30-kDa nominal MW limit (NMWL; Millipore, Bedford, Mass.) for GSH-FR assays. Mold-phase culture supernatants were concentrated and diafiltered with Centricon Plus 20, 10-kDa NMWL, or Centriplus, 30-kDa NMWL (Millipore). To determine the coproduction of siderophores and ferric reducing agents, Ultrafree-15, 5-kDa NMWL, was used. Low-MW fractions were concentrated 10-fold by lyophilization.

Ferric reduction assays. GSH-FR activity was assayed by quantitation of Fe(II) with the chromogenic Fe(II) chelator ferrozine (Sigma) in a microtiter plate-based assay containing 0.5 mg of GSH (Sigma) ml¹ similar to our protocol described previously except buffered to pH 7.0 with 25 mM sodium phosphate (46). Physiological substrates or FeCl₃ was added to final concentrations of 82 µM (41 µM for Fe₂RA₃, 0.89 µM for Fe₂DA₃) and incubated at 37°C for 1 h (15 min for Fe₂DA₃) on a gyratory platform shaker before measuring the A₅₆₂. When FeCl₃ was the sole substrate assayed, the reaction mixture was incubated at 37°C for 2 h. For hemin and PIX, the background A₅₆₂ of the mixture of supernatant, substrate, and ferrozine in the absence of GSH was also subtracted due to the _max of PIX (30). Low-MW ferric reductant and cell surface ferric reducing agent activity were assayed with the chromogenic Fe(II) chelator bathophenanthroline disulfonate as described previously (46). For all assays, samples were compared to a reference blank containing all reaction components with the appropriate buffer as a sample. Each absorbance value minus its reference blank absorbance value was compared to a standard curve generated with FeSO₄, and the amount of Fe(II) produced was normalized to the culture density (cells, low-MW culture supernatant) or to the amount of protein (high-MW culture supernatant) as determined by Non-Interfering Protein Assay (Geno-Tech, St. Louis, Mo.). All assays were performed in triplicate.

Siderophore assays. Siderophores were detected with ferric perchlorate or Microbacterium flavescens JG-9 (ATCC 25091) bioassay essentially as described previously (37). Equal amounts of low-MW culture supernatant or uninoculated media and 5 mM Fe(C10₄)₃ were incubated at 37°C for 8 h, after which the A₄₉₅ was measured. Each absorbance value minus its reference blank absorbance value was compared to a standard curve generated with RA. Uninoculated media, H₂O, FeSO₄, and FeCl₃ were assayed as negative controls for the bioassay, and RA was assayed as a positive control.

TBE-urea-PAGE. Tf isoforms (iron free, N-terminal monoferric Tf, C-terminal monoferric Tf, and diferric Tf) were separated by TBE-urea-PAGE as described previously (28). The gel consisted of 6 M urea-6% acrylamide in TBE (100 mM Tris, 10 mM boric acid, 8 mM EDTA). Samples were mixed with an equal volume of 2× loading buffer (0.4% bromophenol blue, 30% glycerol, 200 mM Tris, 20 mM boric acid, 16 mM EDTA) and electrophoresed in TBE at 100 V for 18 h in a Hoeffer Sturdier SE400 apparatus. Gels were stained by boiling in 0.05% Coomassie brilliant blue R250 (Fisher)-25% isopropyl alcohol-10% glacial acetic acid as described previously (48).

Iron removal from Tf. Siderophore-mediated Fe(III) removal from Tf was assayed by TBE-urea-PAGE similar to a method described previously (14, 44). RA was mixed with Tf (25, 50, or 100% saturated) in a molar ratio of 3:1 (RA/Tf) or 30:1 and incubated at 37°C for 24 h. Df in a molar ratio of 20:1 (Df/Tf) and H₂O were incubated as positive and negative controls, respectively.

Statistical analysis. Statistical significance was determined by using Student's t test.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Relationship between siderophores and ferric reducing agents. Independent production of siderophores and ferric reducing agents by H. capsulatum has previously been described (7, 8, 46). Under conditions of stringent iron limitation, siderophores and GSH-FR were produced in the same culture (Fig. 1), as well as cell surface ferric reducing agent [4.13 ± 0.66 nmol of Fe(II)/10⁹ CFU] and extracellular ferric reductant [7.81 ± 0.72 nmol of Fe(II)/10⁷ CFU]. Siderophores were detected in the iron-limiting culture by use of ferric perchlorate and stimulation of M. flavescens JG-9 growth (data not shown). Ferric reducing activities were higher after growth in iron-limiting conditions (6.3-fold for GSH-FR, 1.5-fold for cell surface ferric reducing agent, and 3.9-fold for ferric reductant) but were detectable after growth in iron-replete conditions [Fig. 1, 2.84 ± 0.07 nmol of Fe(II)/10⁹ CFU for cell surface ferric reducing agent, 2.02 ± 0.06 nmol of Fe(II)/10⁷ CFU for extracellular ferric reductant]. In contrast, after growth under iron-replete conditions, siderophore activity was not detected with ferric perchlorate and M. flavescens JG-9 growth was not stimulated (data not shown). These regulatory patterns are consistent with previous results of individual expression of siderophores or reducing agents (20, 46).

View larger version (8K): [in this window] [in a new window]   FIG. 1.   Production of siderophores (A) and GSH-dependent ferric reductase (B) in single iron-limiting or iron-replete cultures. Siderophore activity was assayed with ferric perchlorate; ferric reductase activity was assayed with the chromogenic Fe(II) chelator ferrozine, GSH, and FeCl3. The averages of triplicate wells from a representative experiment are shown; standard deviations are indicated by bars. Similar results were obtained in three independent experiments.

View larger version (13K): [in this window] [in a new window]   FIG. 2.   GSH-dependent ferric reductase activity with the RA-type hydroxamate siderophores ferric dimerum acid, Fe2RA3, or iron-free RA as substrates. Ferric reduction was assayed with ferrozine, GSH, and the indicated siderophore. The averages of triplicate wells from a representative experiment are shown; standard deviations are indicated by bars. Significantly higher ferric reductase activity was detected with Fe2RA3 than with iron-free RA (P < 0.001). Similar results were obtained in three independent experiments.

Ability of RA to remove Fe(III) from Tf. TBE-urea-PAGE, which separates the iron-free, monoferric and diferric isoforms of Tf, was used to monitor RA-mediated Fe(III) removal from 25, 50, or 100% saturated Tf (holo Tf). After 24 h of incubation at 37°C, the band corresponding to diferric Tf was replaced by bands corresponding to apo Tf and monoferric Tf, in an RA concentration-dependent manner, indicating that iron was removed from diferric Tf (Fig. 3). RA-mediated Fe(III) removal was also observed with 25 and 50% saturated Tf. Siderophore-mediated Fe(III) removal required incubation prior to electrophoresis. Siderophore-mediated Fe(III) removal was also observed with Df, which has a higher affinity for Fe(III) than does Tf (2). Neither degradation nor nonspecific Fe(III) removal occurred during incubation with H₂O.

View larger version (51K): [in this window] [in a new window]   FIG. 3.   TBE-urea-PAGE analysis of 25% saturated Tf (lanes 5 to 8), 50% saturated Tf (lanes 9 to 12), or holo Tf (100% saturated; lanes 13 to 16) incubated with RA in a 3:1 (siderophore/Tf) molar ratio (lanes 5, 9, and 13) or a 30:1 molar ratio (lanes 6, 10, and 14) or Df in a 20:1 molar ratio (lanes 7, 11, and 15). Untreated apo Tf (lane 1), 25% (lane 2) and 50% (lane 3) saturated Tf, and holo Tf (lane 4) were electrophoresed for reference. Tf incubated with H2O (lanes 8, 12, and 16) and the unincubated mixture of RA and 25% (lane 17) and 50% (lane 18) saturated Tf and holo Tf (lane 19) and Df and holo Tf (lane 20) were electrophoresed as controls.

Iron acquisition from host ferric compounds. We examined the ability of host ferric binding compounds to enhance yeast-phase H. capsulatum growth in vitro. Growth rate and culture turbidity were increased when hemin, holo Tf, or FeCl₃ was added to the culture medium (Fig. 4A and B). Partially saturated Tf initially inhibited H. capsulatum growth; however, this effect was reversed at later growth phases. No significant difference was observed when PIX, the Fe(III)-free precursor of hemin, was added; thus, growth enhancement is due to the ability of H. capsulatum to utilize iron from hemin, not the porphyrin. Growth rate and culture turbidity were decreased after the addition of apo Tf, indicating the ability of this compound to bind and sequester contaminating iron present in mRPMI. The buffering capacity of 25 mM HEPES-1 mM NaHCO₃ maintained the culture pH at 7.3 to 7.5 throughout the growth curve. At this pH, Tf can be fully saturated and there is no low pH-mediated release of Fe(III). Therefore, H. capsulatum utilizes Fe(III) bound to Tf as an iron source.

View larger version (25K): [in this window] [in a new window]   FIG. 4.   Effects of the host iron-binding compounds Tf (A) or hemin (B) on H. capsulatum growth in vitro. Log-phase cells were partially depleted of intracellular iron, washed with PBS, and inoculated at a concentration of 2 × 106 to 3 × 106 CFU ml1 in the indicated medium. Growth was measured by culture turbidity. Supplements added to the basal medium are indicated by the following symbols: for panel A, , no supplement; , 1 µM apo Tf; , 1 µM 25% saturated Tf; , 1 µM 50% saturated Tf; , 1 µM holo Tf; , 2 µM FeCl3; and for panel B, , no supplement; , 1 µM hemin; , 1 µM PIX; , 1 µM FeCl3. The averages of duplicate cultures from a representative experiment are shown; standard deviations are indicated by bars. Similar results were obtained in three independent experiments. (C) GSH-dependent ferric reductase activity with host ferric compounds as substrates. Ferric reduction was assayed with ferrozine, GSH, and the indicated compound. The averages of triplicate wells from a representative experiment are shown; standard deviations are indicated by bars. The amount of reductase activity was significantly higher on the substrate of hemin than PIX (P < 0.0005) and on holo Tf than 50% saturated Tf (P < 0.01), 25% saturated Tf (P < 0.001), or apo Tf (P < 0.0001). Similar results were obtained in three independent experiments.

Iron acquisition from environmental ferric compounds. We examined the effects of Streptomyces Df (iron free), FOB (iron replete, or Fe-Df), and the fungal siderophore FC (iron free) and Fe-FC (iron replete) on yeast- and mold-phase growth in vitro. The addition of FOB or Fe-FC resulted in increased yeast growth rates and culture turbidities to levels similar to those seen with equimolar FeCl₃ (Fig. 5A). The addition of iron-free FC or Df resulted in decreased culture turbidities and growth rates, presumably due to chelation of contaminating iron. Df demonstrated greater growth-limiting ability than FC. We also assayed the effect of foreign siderophores on mold-phase growth. Yeasts were plated to facilitate quantitation, followed by incubation at room temperature for mold growth; the efficiency of phase conversion and colony growth is low under these conditions. The number of mold colonies (average of two plates) was increased compared to when no supplement was added from 6 to >90, >150, and 24 after being made iron replete by the addition of FOB, Fe-FC, or FeCl₃, respectively; 8 and 32 colonies formed on medium supplemented with Df or FC, respectively.

View larger version (10K): [in this window] [in a new window]   FIG. 5.   (A) Effect of foreign siderophores on H. capsulatum yeast growth in vitro. Log-phase cells were partially depleted of intracellular iron, washed with PBS, and inoculated at a concentration of 2 × 106 to 3 × 106 CFU ml1 in the indicated medium. Growth was measured by culture turbidity. Supplements added to the basal medium are indicated by the following symbols: , no supplement; , 5 µM Df; , 5 µM FOB; , 5 µM FC; , 5 µM Fe-FC; , 5 µM FeCl3. The averages of duplicate cultures from a representative experiment are shown; standard deviations are indicated by bars. Similar results were obtained in three independent experiments. (B) GSH-dependent ferric reductase activity with foreign siderophores as substrates. Ferric reduction was assayed with ferrozine, GSH, and the indicated compound. The averages of triplicate wells from a representative experiment are shown; standard deviations are indicated by bars. The low amount of reduction observed with Df is below the lower detection limit of the spectrophotometer; the level of ferric reduction is significantly higher with FOB as a substrate (P < 0.01). Similar results were obtained in three independent experiments.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Free iron is scarce in the environment and the host. Instead, iron in soil is present as insoluble ferric hydroxides or ferric chelates such as foreign ferric siderophores. In the host, iron is tightly bound to compounds such as holo Tf or hemin. These large macromolecules cannot diffuse into the cell and thus are not readily available to H. capsulatum. One mechanism by which these ferric compounds may be utilized involves the reduction of Fe(III) chelated to the macromolecule releasing free Fe(II). The import of free Fe(II), free Fe(III), or native ferric siderophores has not been described in H. capsulatum but may be similar to the uptake mechanisms for Fe(II) or ferric siderophores characterized in other fungi. The ability of the GSH-FR to reduce Fe(III) chelated to environmental and host compounds correlates with the ability of these compounds to enhance H. capsulatum growth in vitro.

We have examined interactions between different potential iron acquisition systems of H. capsulatum and the utilization of both environmental and host iron sources by the extracellular GSH-FR. In the soil, GSH-FR could facilitate iron acquisition through the production of unchelated soluble Fe(II) from foreign ferric siderophores. Siderophores are present in physiologically high levels in soil, with FOB at 0.1 to 10 µM (24). We have demonstrated a correlation between the ability of the Streptomyces siderophore FOB to enhance H. capsulatum growth in mold and yeast phases and the ability of this compound to serve as a substrate for the GSH-FR. Iron chelated to the fungal siderophore FC is not reduced by the GSH-FR, although Fe-FC does enhance yeast- and mold-phase H. capsulatum growth. This iron is presumably used by some other mechanism, such as receptor-mediated uptake, reduction by cell surface ferric reducing agent or extracellular ferric reductant, or iron competition by H. capsulatum siderophores.

Both reductive and nonreductive mechanisms mediate siderophore uptake for other microbes. In the phytopathogenic fungus Ustilago maydis, a cell surface ferric reducing agent mediates the utilization of FOB, but native Fe-FC is not a substrate for the reducing agent and is imported without ferric reduction. Fe(III) chelated to native ferrichrome A, however, is transported reductively into Ustilago sphaerogena cells (3, 12). The zygomycete Rhizopus microsporus, which causes mucormycosis in patients treated with Df, utilizes FOB by reductive, energy-dependent uptake that does not require internalization of the siderophore (9). The non-siderophore-producing soil yeast Saccharomyces cerevisiae can grow with FOB, FC, triacetylfusarinine C, enterobactin, or RA as iron sources. It utilizes foreign siderophores both by high-affinity, energy-dependent transporters for triacetylfusarinine C, enterobactin, FC, and FOB and by reduction of these siderophores and Fe₂RA₃ by the plasma membrane ferric reductases Fre1p, Fre2p, and Fre3p. Reductive uptake is induced by less-stringent iron limitation than nonreductive transport (15-17, 24, 25, 50).

Although native ferric siderophore uptake has not been described in H. capsulatum, uptake of Fe(III) chelated to DA, coprogen, triacetylfusarinine, and RA has been described in Stemphylium botryosum, Mycelia sterilia, and Rhodotorula pilimanae, which also produce these siderophores but have not been reported to reduce Fe(III) extracellularly. DA and coprogen are imported by active transport in S. botryosum, and the entire Fe(III)-coprogen complex is imported (29). In M. sterilia, triacetylfusarinine removes Fe(III) from RA and is imported by a nonreductive mechanism (1). RA remains extracellular as Fe(III) is imported by nonreductive active transport in R. pilimanae (32).

Import mechanisms for free Fe(II) or Fe(III) have not been described in H. capsulatum. A conserved high-affinity Fe(II) uptake system has been characterized in S. cerevisiae, Schizosaccharomyces pombe, and Candida albicans. Fe(II) produced by cell surface NADPH-dependent ferric reductases is oxidized by a membrane-bound oxidase and is imported by a permease (4, 23, 31, 38, 40, 43). Such high-affinity Fe(II) uptake is regulated by iron and, in the case of C. albicans, is required for virulence. High-affinity iron uptake in Cryptococcus neoformans requires energy and is specific for Fe(II) produced by a cell surface NADPH-dependent ferric reductase and extracellular 3-hydroxyanthraniline (21, 36). Other S. cerevisiae cell surface transporters have low affinity for Fe(II) and recognize multiple transition metals (26). Similar high- or low-affinity systems may import Fe(II) into H. capsulatum.

In the host, one possible iron source in the intracellular compartment occupied by H. capsulatum is holo Tf. Although the host iron sources accessible to intracellular H. capsulatum have not been completely characterized, the intracellular compartment occupied by H. capsulatum in RAW264.7 macrophages contains ferritin and thus has undergone endosomal fusion (45). We have shown that H. capsulatum can grow with holo Tf as an iron source at pH 7.3 to 7.5; that Fe(III) in holo Tf can be reduced to unbound, soluble Fe(II) by the GSH-FR at pH 7.0; and that RA can remove Fe(III) from Tf. Thus, both reduction by the GSH-FR and chelation by siderophores may facilitate utilization of holo Tf in the absence of acid pH-mediated Fe(III) release.

The reductive utilization of Fe(III) bound by mammalian compounds has been described for prokaryotic and eukaryotic cells. The intracellular bacterial pathogen Listeria monocytogenes produces a cell surface ferric reductase that can utilize 50% saturated Tf and holo Tf as substrates (10, 11). The oral spirochete Treponema denticola produces both hemin-binding proteins and a cell surface ferric reducing agent but does not transport hemin into the cytoplasm (42). HeLa S3 whole cells reduce holo Tf and ferric ammonium citrate at pH 7.4; apo Tf inhibits both reactions (27).

Acquisition of iron in the host has been shown to be crucial for H. capsulatum virulence, and specific sequestration of iron is a demonstrated host defense against H. capsulatum. H. capsulatum may acquire iron both through chelation and/or reduction of Fe(III); additionally, Fe(III) would be released if the phagolysosomal pH reached 6.0 (Fig. 6). We have described a potential relationship between these iron acquisition mechanisms and a role for the extracellular GSH-FR in the utilization of host ferric compounds. The structurally similar siderophore RA, which is also a substrate for the GSH-FR, can chelate Fe(III) bound to Tf. Siderophore-bound iron may be then imported. If the pH of the H. capsulatum phagolysosome reached 6.0, then free Fe(III) would be released from holo Tf and may be imported. Additionally, native ferric siderophores, host iron-binding compounds such as holo Tf and hemin, and free Fe(III) are reduced by the GSH-FR. Free, soluble Fe(II) is released and may then be imported. Thus, the different iron acquisition mechanisms could be utilized interactively. Alternately, these potential iron acquisition mechanisms may be dominant under different microenvironmental conditions for mold in soil or yeast in a mammalian host. The production of both siderophores and ferric reducing agents in a single microbial culture is novel to our knowledge. Although both types of iron acquisition mechanisms show increased activity after growth under iron-limiting conditions, the regulation of siderophores and ferric reducing agents differs under iron-replete conditions. This expression difference would be consistent with different roles for siderophores and ferric reductase. The identification of regulatory, biosynthetic, or structural genes for siderophore and ferric reductase production will enable expression studies and mutational analyses to elucidate the biological and pathogenic roles of these mechanisms.

View larger version (25K): [in this window] [in a new window]   FIG. 6.   Model of iron acquisition during H. capsulatum macrophage infection.