Leishmania-Mediated Inhibition of Iron Export Promotes Parasite Replication in Macrophages

Leishmania parasites infect macrophages, cells that play an important role in organismal iron homeostasis. By expressing ferroportin, a membrane protein specialized in iron export, macrophages release iron stored intracellularly into the circulation. Iron is essential for the intracellular replication of Leishmania, but how the parasites compete with the iron export function of their host cell is unknown. Here, we show that infection with Leishmania amazonensis inhibits ferroportin expression in macrophages. In a TLR4-dependent manner, infected macrophages upregulated transcription of hepcidin, a peptide hormone that triggers ferroportin degradation. Parasite replication was inhibited in hepcidin-deficient macrophages and in wild type macrophages overexpressing mutant ferroportin that is resistant to hepcidin-induced degradation. Conversely, intracellular growth was enhanced by exogenously added hepcidin, or by expression of dominant-negative ferroportin. Importantly, dominant-negative ferroportin and macrophages from flatiron mice, a mouse model for human type IV hereditary hemochromatosis, restored the infectivity of mutant parasite strains defective in iron acquisition. Thus, inhibition of ferroportin expression is a specific strategy used by L. amazonensis to inhibit iron export and promote their own intracellular growth.


Introduction
Intracellular parasites must obtain essential nutrients from their host cells. Iron is a vital nutritional resource for mammalian hosts and also for many pathogens, acting as an essential cofactor of proteins and enzymes involved in metabolic pathways. Under physiological conditions, iron is normally found in the insoluble, oxidized Fe 3+ form associated with carrier proteins such as transferrin and ferritin. After endocytosis mediated by transferrin receptors, Fe 3+ is reduced to the soluble Fe 2+ form and translocated to the cytosol. In the cytosol, iron may be utilized by the host, stored as complex with ferritin, or exported out of the cells. Thus, the pool of available iron within mammalian cells is determined by a carefully orchestrated balance between uptake through plasma membrane and endosomal receptors/transporters and cellular export.
The only known mammalian iron exporter is ferroportin (Fpn1), a protein with an essential role in iron homeostasis [1]. Fpn1 is expressed on the surface of cells specialized in storage or transport of iron, including enterocytes, hepatocytes and macrophages. Transcription of Fpn1 is increased by several factors, including iron. Further, Fpn1 mRNA has a 59iron-responsive element (IRE) that limits its translation when iron availability is low and enhances it under high iron conditions. Fpn1 levels on the cell surface are regulated by hepcidin, a peptide hormone that triggers Fpn1 internalization and lysosomal degradation [2]. Modulation of Fpn1 expression allows macrophages to function as mediators of iron homeostasis in vivo, responding to systemic and localized signals by retaining or exporting iron [3,4]. Missense mutations in Fpn1 are autosomal dominant in humans and some mutations lead to hereditary hemochromatosis type IV (form A), a human disease characterized by high ferritin levels, low transferrin saturation and iron accumulation inside macrophages [5]. Other Fpn1 mutations lead to constitutive iron export due to the inability to respond to hepcidin.
The protozoan parasite Leishmania, which causes serious infections in millions of people throughout the world, replicates mainly in macrophages. Until recently very little was known about mechanisms of iron acquisition and utilization by Leishmania, and how changes in host cell iron homeostasis affect these intracellular parasites. Recent studies improved this scenario, by identifying three L. amazonensis membrane proteins that mediate iron uptake and are required for parasite virulence: the Fe 3+ reductase LFR1 [6], the Fe 2+ transporter LIT1 [7] and the heme transporter LHR1 [8]. In L. donovani, the ATP-binding cassette membrane protein LABCG5 was also implicated in the salvage of heme generated by degradation of hemoglobin in parasite lysosomes [9]. These studies showed that the iron-poor environment of macrophage phagolysosomes induces Leishmania to express molecules that promote uptake of iron and heme, factors essential for intracellular survival and replication. However, it is still not understood how the parasites compete with the highly developed pathways of iron export and storage of macrophages to gain access to iron. In this study we examined how iron efflux and modulation of host cell iron content affect the intracellular growth of L. amazonensis. Our results revealed a parasite strategy for inhibition of iron export that is required for the establishment of productive infections in macrophages.

Infection with L. amazonensis inhibits macrophage expression of Fpn1 in a TLR4-dependent manner
We investigated the impact of infection with L. amazonensis on the expression of Fpn1 in mouse bone marrow-derived macrophages (BMDM). Low levels of Fpn1 transcripts were detected in wild type and TLR4 2/2 BMDM with and without L. amazonensis infection ( Figure 1A,B). In agreement with previous studies showing Fpn1 upregulation in response to iron loading [10,11], treatment with Fe-nitrilotriacetate (Fe-NTA) for 16 h markedly increased Fpn1 transcripts ( Figure 1A,B). Following infection with L. amazonensis, Fpn1 mRNA levels were reduced in Fe-NTA-treated wild type BMDM ( Figure 1A). In contrast, although TLR4 2/2 BMDM showed higher amounts of Fpn1 mRNA (particularly 24 h after Fe-NTA treatment), the transcript levels were very similar between infected and non-infected cells ( Figure 1B). At the protein level, immunoblots detected low amounts of Fpn1 in untreated BMDM and elevated levels 48 h after Fe-NTA treatment, in noninfected BMDM derived from either wild type or TLR4 2/2 mice ( Figure 1C). Similar to what was observed with the mRNA, infection with L. amazonensis reduced the amounts of Fpn1 protein in Fe-NTA-treated wild type, but not TLR4 2/2 BMDM. Fpn1 levels detected in BMDM not treated with Fe-NTA were low, but quantification and ratio determinations in relation to actin showed a similar trend under these conditions as well -Fpn1was reduced in infected wild type, but not in infected TLR4 2/2 BMDM ( Figure 1C). Similar experiments performed with BMDM from TLR2 2/2 mice showed results undistinguishable from wild type (data not shown). These findings suggest that the signalling pathway involved in Fpn1 downregulation by L. amazonensis is TLR4-dependent.

L. amazonensis infection induces TLR4-dependent expression of hepcidin in macrophages
The peptide hormone hepcidin is a potent regulator of iron export in macrophages, inducing Fpn1 internalization and degradation [2,12]. We found that L. amazonensis infection upregulates hepcidin (Hamp) transcripts in BMDM (Figure 2A), an effect not observed with heat-killed parasites (data not shown). The ability of L. amazonensis to induce hepcidin expression was TLR4-dependent, consistent with what was observed for Fpn1 downregulation ( Figure 1C). Importantly, upregulation of Hamp transcripts was also observed in vivo, in the liver and in footpad tissue associated with cutaneous lesions of mice infected for 9 weeks with L. amazonensis ( Figure 2B,C) (see also supporting information S1). These results strongly suggest that macrophages infected with L. amazonensis express hepcidin.

L. amazonensis infection induces iron accumulation and ferritin upregulation in macrophages
The inhibition in Fpn1 surface expression induced by L. amazonensis infection was expected to increase the intracellular iron content of macrophages. To test this hypothesis, we used a ferrozine assay to quantify iron in BMDM infected or not with L. amazonensis amastigotes. The results showed a significant increase in the total iron content of wild type BMDM at 24 and 48 h after infection. This effect was not seen in TLR4 2/2 BMDM (Table 1). At the same time points, infected wild type BMDM showed upregulation of the iron storage protein ferritin, at both the mRNA and protein levels ( Figure 3A, B). This increase in intracellular iron content is likely to be a consequence of the parasite-induced inhibition of Fpn1 expression, which is expected to result in decreased iron export.
Hepcidin-mediated downregulation of Fpn1 stimulates the intracellular growth of L. amazonensis We proceeded to investigate the impact of Fpn1-mediated iron export on the intracellular replication of L. amazonensis. In untreated BMDM the parasites increased in number over time, doubling in density by 48 h after infection. When BMDM were pre-treated with 1 mg/ml hepcidin for 4 h before infection, the intracellular growth rate of the parasites was enhanced ( Figure 4A,B). These results suggest that hepcidin-mediated Fpn1 downregulation and block in iron export is beneficial for the establishment of L. amazonensis intracellular infections. After pre-treating BMDM with Fe-NTA, a condition that markedly stimulates Fpn1 expression (Figure 1), the opposite effect was observed: L. amazonensis intracellular replication was inhibited. This inhibitory effect of iron loading was apparently not due to iron toxicity, since Fe-NTA-treated BMDM remained .80% viable up to 48 h (Alamar Blue viability assay, results not shown). Furthermore, the inhibition in parasite replication was fully reversed by a subsequent exposure of the Fe-NTA-treated BMDM to hepcidin ( Figure 4A,B). Microscopic examination showed that Fe-NTA-loaded macrophages contained small parasitophorous vacuoles (PV) with only one or two amastigotes. In contrast, after hepcidin treatment PVs were markedly expanded and housed numerous replicating parasites ( Figure 4B). These findings suggest that even when macrophages are loaded with an iron source such as Fe-NTA, continuous iron efflux through Fpn1 on the plasma membrane has a negative impact on the ability of L. amazonensis to

Author Summary
Infection with the protozoan parasite Leishmania causes significant human disease in many parts of the world, particularly in the Middle East, India and South America. The parasite is transmitted by sand flies, which are difficult to control and are becoming increasingly common in urban areas. With domestic dogs serving as reservoirs of the disease and global travel increasing the population of infected human patients, the overall burden of leishmaniasis is on the rise. In mammals these parasites replicate inside macrophages, and therefore need strategies to survive within a cell that is specialized in killing pathogens. Earlier work demonstrated that iron is one of the essential nutrients that Leishmania must acquire from host cells to survive. Acquiring iron is particularly challenging inside macrophages, which play an important role in host iron homeostasis and export iron extracellularly through the membrane transporter ferroportin. We found that Leishmania amazonesis induces their host macrophages to produce hepcidin, a peptide that triggers internalization and degradation of ferroportin. This strategy increases the macrophage intracellular iron pool, and stimulates Leishmania replication. These results suggest that defects in iron homeostasis, which occur frequently in the human population, can have an important role in susceptibility to Leishmania infections.
replicate in macrophages. Confirming the importance of hepcidin production by macrophages for the establishment of intracellular infections, the growth rate of L. amazonensis was reduced in BMDM prepared from Hamp 2/2 mice, when compared to their wild type counterparts ( Figure 4C).

Modulation of plasma membrane-associated Fpn1 controls the intracellular replication of L. amazonensis
Our results suggested that hepcidin production during L. amazonensis infection reduces the amount of Fpn1 present on the surface of macrophages, a condition that promotes intracellular parasite replication. In order to directly visualize this process, we transfected BMDM with three different GFP-tagged constructs: Wild Type Fpn1, Fpn1(H32R) and Fpn1(N144H). Fpn1(H32R) has a dominant-negative effect, severely reducing the amount of Fpn1 that reaches the cell surface [13]. On the other hand, the N144H mutation renders Fpn1 resistant to hepcidinmediated internalization and degradation, increasing levels at the cell surface [14]. As expected, we detected Fpn1-GFP on the plasma membrane of BMDM transfected with the wild type and hepcidin-resistant N144H constructs, and on intracellular compartments of BMDM expressing the dominant-negative H32R mutation ( Figure 5, Control). Hepcidin treatment triggered removal of wild type Fpn1-GFP from the plasma membrane of macrophages, while the hepcidin-resistant Fpn1(N144H)-GFP These results further support the hypothesis that L. amazonensis downregulates Fpn1 expression by triggering endogenous hepcidin production by macrophages. The hepcidin-resistant version of Fpn1-GFP remained associated with the BMDM plasma membrane up to 72 h after infection (data not shown).
Intracellular replication assays showed that L. amazonensis growth in BMDM overexpressing wild type Fpn1-GFP was inhibited, in agreement with the increased levels of surfaceassociated Fpn1 and lower levels of cellular iron. In contrast, expression of the dominant-negative Fpn1(H32R)-GFP construct that interferes with Fpn1 surface targeting stimulated parasite intracellular replication. Reinforcing the conclusion that intracellular growth of L. amazonensis depends on hepcidin-mediated removal of Fpn1 from the macrophage surface, parasite replication over a period of 48 h was strongly inhibited in BMDM expressing the hepcidin-resistant Fpn1(N144H)-GFP ( Figure 6).

Dominant-negative Fpn1 rescues the intracellular growth of Leishmania mutants defective in iron acquisition
We hypothesized that the enhanced replication of wild type L. amazonensis in BMDM expressing dominant-negative Fpn1(H32R) was due to an increased availability of intracellular iron to the parasites. To test this hypothesis, we examined the effect of dominant-negative Fpn1 on the intracellular fate of L. amazonensis null mutant strains deficient in iron uptake. Strains lacking the LIT1 ferrous iron transporter (Dlit1) and the LFR1 ferric iron reductase (Dlfr1) are incapable of growing inside macrophages, but their replication can be fully rescued by loading the endosomal compartment of host macrophages with an abundant iron source, cationic ferritin [6]. Strikingly, we found that expression of dominant-negative Fpn1 restored intracellular growth rates of Dlit1 and Dlfr1 null strains to levels comparable to wild type L. amazonensis ( Figure 7A, B). A similar effect was observed in BMDM from flatiron mice [13], which carry the dominantnegative H32R mutation in one Fpn1 allele, and have elevated levels of ferritin ( Figure 8A) and iron ( Figure 8B). The Dlit1 and Dlfr1 null parasite strains did not grow intracellularly in control C3H BMDM, but were able to replicate in ffe/+ BMDM ( Figure 8C,D). Collectively, these results show that Fpn1 removal from the cell surface facilitates L. amazonensis access to iron while replicating inside macrophage PVs, and compensates for specific defects in the parasite's iron acquisition pathway.

Discussion
Genes involved in iron acquisition play a critical role in pathogen virulence, as extensively demonstrated in bacteria [15] and more recently in the protozoan parasite Leishmania amazonensis [6,7]. In mammals Leishmania is an obligate intracellular parasite of macrophages, replicating within PVs with properties of phagolysosomes. Utilization of the macrophage as host cell has important implications for how Leishmania gains access to iron. Macrophages from the reticuloendothelial system play a fundamental role in iron recycling in vivo, through the process of erythrophagocytosis. Heme released from phagocytosed senescent red blood cells is translocated to the macrophage cytosol, where iron is extracted through the activity of heme oxygenase. Cytosolic iron is then utilized by the cell, stored as a complex with ferritin, or exported by Fpn1 [1]. In this study we showed that L. amazonensis directly interferes with the iron export function of macrophages, by inhibiting cell surface expression of Fpn1. This Leishmania-driven process is associated with increased total macrophage iron content and stimulation of parasite intracellular replication (Figure 9).
Consistent with an inhibitory effect on iron export, infection by L. amazonensis enhanced expression of ferritin, the cytosolic iron storage protein. At the protein level, this observation is consistent with the presence of a 59 IRE in the ferritin mRNA [16]. However, the total amount of ferritin mRNA was also elevated after L. amazonensis infection. This indicates that signaling events regulating transcription, possibly associated with macrophage inflammatory responses, are also triggered during infection and participate in the cellular response to elevated iron pools. Indeed, we found that the ability of Leishmania to inhibit Fpn1 expression depends on TLR4, a major pattern-recognition receptor of macrophages that is commonly associated with pro-inflammatory responses. Different species of Leishmania were described to elicit distinct TLR4-mediated responses in macrophages, although the exact identity of the molecule(s) serving as TLR4 ligand(s) remain unknown [17]. While TLR4 plays a protective role in infections with L. major [18], it appears to have an anti-inflammatory effect in macrophages infected with L. mexicana, a species closely related to the L. amazonensis used in our study [19]. We found that TLR4dependent downregulation of Fpn1 requires live parasites, suggesting that the molecule(s) involved may be actively secreted, or may only come in contact with the TLR4 receptors in an intracellular compartment generated during cell invasion.
Several lines of evidence indicate that L. amazonensis-induced Fpn1 downregulation is mediated, at least in part, by production of the peptide hormone hepcidin by infected macrophages. First, there was a marked increase in hepcidin transcripts in infected macrophages. Second, Fpn1 carrying the N144H mutation that renders it resistant to hepcidin-mediated internalization and degradation [13] remained on the plasma membrane of infected macrophages, while wild type Fpn1 was removed. Third, the effects of L. amazonensis infection on both hepcidin and Fpn1 expression in macrophages were TLR4-dependent. Hepcidin is produced in the liver during inflammation and other stimuli in a paracrine manner [20], but it can also be produced locally by macrophages in an autocrine fashion in response to infections, as previously shown for Pseudomonas aeruginosa [21] and Mycobacterium tuberculosis [22]. Similar to our findings with L. amazonensis, hepcidin production and decrease cell surface-associated Fpn1 triggered by infection with Pseudomonas aeruginosa require TLR4 [21]. Earlier studies demonstrated that Fpn1 containing the H32R mutation has a dominant-negative effect, preventing the traffic of Fpn1 to the cell surface [13]. We found that Fpn1 carrying this mutation enhances L. amazonensis intracellular replication when expressed in macrophages. This was observed after expression of Fpn1(H32R)-GFP, and also in macrophages from flatiron mice, which are heterozygous for this mutation and represent a model for the autosomal-dominant human type IV hereditary hemochromatosis type A [13]. Notably, macrophages expressing Fpn1(H32R)-GFP or isolated from flatiron mice rescued the intracellular growth defect of LFR1 and LIT1 null mutants deficient in iron uptake [6,7], directly supporting the conclusion that Fpn1 removal from the cell surface facilitates parasite access to iron.
Our findings indicate that blocking Fpn1 surface expression increases iron availability in the cytosol, which is associated with L. amazonensis intracellular replication. This result suggests the existence of a crosstalk between the cytosolic iron pool and the iron content of L. amazonensis PVs, a process that is still poorly understood ( Figure 9). Interestingly, we found that increasing ferritin-iron cytosolic stores by treating macrophages with an exogenous iron source is not sufficient to promote parasite replication. Hepcidin-mediated downregulation of Fpn1 appears to be required to facilitate amastigote access to iron, suggesting that Fpn1 internalization promotes formation of a cytosolic iron pool that can be more readily mobilized by the parasites. Evidence for such a mechanism was recently obtained in breast cancer cells, where hepcidin-mediated internalization of Fpn1 was associated with an increased labile iron pool [23]. Another intriguing possibility is that Fpn1 may be delivered to the membrane of the lysosome-like PV after hepcidin-mediated internalization, where it may remain transiently active and transport cytosolic iron into the lumen of the parasite-containing vacuole.
Fpn1-mediated iron depletion inhibits the intracellular growth of several bacterial pathogens, such as Chlamydia psittaci, Chlamydia trachomatis, Legionella pneumophila [24]. Elevated Fpn1 expression was also identified as a host defense mechanism against infections with Salmonella typhimurium and Mycobacterium tuberculosis [25]. To our knowledge, the present study represents the first demonstration that a protozoan pathogen blocks Fpn1 expression as a strategy to facilitate their intracellular replication. This strategy may complement additional mechanisms that facilitate parasite access to iron, such as the elevated expression of transferrin receptors observed in macrophages infected with Leishmania donovani [26]. Our study demonstrates the important role played by cytosolic iron pools in L. amazonensis infections, and opens the way for future investigations of how iron enters the PV, and whether differential access to iron in cutaneous or visceral sites influences the clinical disease caused by different Leishmania species.  heat-inactivated FBS, 5% penicillin/streptomycin, 0.1% hemin (from a 25 mg/ml stock in 50% triethanolamine), 10 mm adenine, and 5 mm l-glutamine. To obtain axenic amastigotes, stationary phase cultures rich in metacyclic promastigotes were incubated at 32uC in acidified media, as previously described [6]. Heat-killed (65uC incubation for 1 h) axenic amastigotes were washed, resuspended in BMDM medium and added to BMDM at a MOI = 5.

BMDM treatment and infection
BMDM were plated on glass coverslips in RPMI 10% FBS. After 24 h, BMDM were incubated with freshly prepared Fenitrilotriacetate (Fe-NTA, 1:4 solution of FeCl3 200 mM and NTA 400 mM) for 16 h, and washed before infection. For infection, axenic amastigotes were added to BMDM for 1 h at 34uC at a MOI = 5, washed in PBS to remove non-internalized parasites, reincubated for 1, 24, 48 or 72 h at 34uC, fixed with 4% PFA, stained with DAPI and anti-L. amazonensis polyclonal mouse antibodies, and mounted (Prolong Antifade; Molecular Probes). Intracellular parasites were quantified microscopically, as previously described [29]. To assess the effect of hepcidin on inhibition of ferroportin expression and parasite growth, BMDM were treated for 4 or 18 h with 1 mg/ml hepcidin (kindly provided by J. Kaplan and D. M. Ward, University of Utah) [29].

Imaging
For live imaging, transfected BMDM were plated in 35 mm glass-bottom dishes (Mattek), infected or not with axenic amastigotes for 1 h, washed and reincubated for 48 h. Dishes were then transferred to an environmental chamber at 34uC with 5% CO 2 and humidity control (Pathology Devices). Confocal   week-old female C57BL/6 mice were inoculated with 1610 6 purified metacyclics of wild type L. amazonensis (Pinto-da-Silva et al., 2005) in the left hind footpad in a volume of 50 ml PBS. After 9 weeks, liver or footpad from 3 infected mice were recovered and vigorously homogenized using micro tissue grinder (Radnoti LLC). Non-infected mice were used as experiment controls. The samples were kept on ice for subsequent RNA extraction. cDNA synthesis was carried out using 1 mg of total RNA and Superscript II Reverse Transcriptase (Invitrogen) according to the manufacturer's protocol. To quantify transcript levels, 1 ml of cDNA was amplified using the following primers: mouse FPN1-Forward 59-CTA CCATTAGAAGGATTGACCAGCT-39, mouse FPN1-Reverse 59-ACTGGAGAACCAAATGTCATAATCTG-39; mouse HAMP-Forward 59-TGTCTCCTGCTTCTCCTCCT-39, mouse HAMP-Reverse59-CTCTGTAGTCTGTCTCAT-CTGTTG-39; mouse FTH1-Forward 59-GCGAGGTGGCC-GAATCT-39, mouse FTH1-Reverse 59-CAGCCCGCTCTCC-CAGT-39; mouse HPRT1-Forward 59-TCAGTCAACGGGG-GAACATAAA-39, mouse HPRT1-Reverse 59-GGGGCT-GTACTGCTTAACCAG-39, mouse b-III-Tubulin-Forward 59-CGCACGACATCTAGGACTGA-39, mouse b-III-Tubulin-Reverse 59-TGAGGCCTCCTCTCACAAGT-39 Real-time PCR was performed using a BioRad iQ icycler Detection system (BioRad Laboratories) using SYBR green fluorophore (BioRad Laboratories) according to the manufacturer's instructions. Fluorescence was detected at each annealing step, and the cycle threshold (C t ) was calculated by determining the point at which the fluorescence exceeded a threshold limit. All reactions were performed in triplicate, and negative controls (no template cDNA) were included in each experiment. Data were normalized to the level of HPRT1 (mouse BMDM) or tubulin (mouse liver and footpad) expression in each sample.

Protein extraction and western blot
BMDM were washed twice with cold PBS, scraped from the dish, collected by centrifugation (335 g) for 10 min at 4uC, lysed in 30 ml RIPA buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.50% sodium deoxycholate, 0.10% SDS) with protease inhibitors (Roche), and centrifuged at 14000 g for 15 min at 4uC to remove insoluble material. Lysates were assayed for protein content (Pierce BSA Protein Assay kit), 60 mg per sample were mixed with SDS sample buffer at room temperature for 30 min, separated by SDS-PAGE (without sample boiling) and transferred to a PVDF membrane (Millipore). Membranes were incubated overnight with 1:1000 rabbit anti-Fpn1 antibody (Alpha Diagnostics), 1:800 anti L-ferritin (Abcam) or 1:5000 anti-actin (Sigma) in blocking buffer (PBS 3% milk, 0.1% Tween), followed by HRP-coupled anti-rabbit IgG or anti mouse IgG secondary Macrophages infected by Leishmania amazonensis produce hepcidin, a peptide hormone that binds to the iron exporter ferroportin, inducing its internalization and degradation. This causes an increase in the macrophage cytosolic iron content, which includes the bioavailable labile iron pool and iron stored as a complex with ferritin. By an unknown mechanism, Fpn1 internalization and elevated cytosolic iron increases iron availability inside Leishmaniacontaining parasitophorous vacuoles, stimulating parasite growth. doi:10.1371/journal.ppat.1003901.g009 antibodies (Amersham Biosciences). Blots were developed using Immun-Star HRP Luminol Enhancer and peroxidase buffer (Bio-Rad) and detected with a Fuji LAS-3000 Imaging System and Image Reader LAS-3000 software. Digital quantifications of chemiluminescence were performed using Image J software.

Intracellular iron quantification
A ferrozine-based colorimetric assay was performed as previously described [30]. Briefly, BMDMs were infected or not with L. amazonensis axenic amastigotes for 1 h and reincubated for 2, 24 or 48 h. At different time points, the dishes were washed twice with ice-cold PBS and stored at 220uC. Cells were lysed with 200 ml of 50 mM NaOH for 2 h on a shaker in a humidified atmosphere and aliquots were used for iron and protein determinations. For iron quantification, 100 ml of lysates were mixed with 100 ml 10 mM HCl, and 100 ml of the iron-releasing reagent (a freshly mixed solution of equal volumes of 1.4 M HCl and 4.5% (w/v) KMnO 4 ). After for 2 h at 60uC in a fume hood, samples were cooled to room temperature and 30 ml of the iron-detection reagent (6.5 mM ferrozine, 6.5 mM neocuproine, 2.5 M ammonium acetate, and 1 M ascorbic acid) was added to each sample. After 30 min, 280 ml of each sample was transferred into a 96-well plate and the absorbance measured at 550 nm in a microplate reader. The iron content in each sample was calculated based on a standard curve (0-300 mM FeCl 3 ) prepared under the same conditions, and normalized to protein levels. Protein levels were determined by a modified Lowry assay (Pierce) per manufacturer's instructions.

Supporting Information
Supporting Information S1 Raw data of qPCR assays detecting hepcidin (Hamp) in mouse tissues. The tables provided show the Ct values obtained in real time PCR assays with RNA isolated from 3 uninfected mice or 3 mice infected with L. amazonensis using specific primers for Hamp or tubulin, as described in Materials and Methods. The table also indicates how normalization was done using the values for tubulin transcripts. (PDF)