Slc11a1 limits intracellular growth of Salmonella enterica sv. Typhimurium by promoting macrophage immune effector functions and impairing bacterial iron acquisition

Abstract

The natural-resistance associated macrophage protein 1, Slc11a1, is a phagolysosomal transporter for protons and divalent ions including iron, that confers host protection against diverse intracellular pathogens including Salmonella. We investigated and compared the regulation of iron homeostasis and immune function in RAW264.7 murine phagocytes stably transfected with non-functional Slc11a1 and functional Slc11a1 controls in response to an infection with Salmonella enterica serovar Typhimurium (S. Typhimurium). We report that macrophages lacking functional Slc11a1 displayed an increased expression of transferrin receptor 1, resulting in enhanced acquisition of transferrin-bound iron. In contrast, cellular iron release mediated via ferroportin 1 was significantly lower in Salmonella-infected Slc11a1-negative macrophages in comparison to phagocytes bearing Slc11a1. Lack of Slc11a1 led to intracellular persistence of S. Typhimurium within macrophages which was paralleled by a reduced formation of nitric oxide, tumour necrosis factor-alpha and interleukin-6 in Slc11a1-negative macrophages following Salmonella infection, whereas interleukin-10 production was increased. Moreover, Slc11a1-negative phagocytes exhibited higher cellular iron content, resulting in increased iron acquisition by intracellular Salmonella. Our observations indicate a bifunctional role for Slc11a1 within phagocytes. Slc11a restricts iron availability, which firstly augments pro-inflammatory macrophage effector functions and secondly concomitantly limits microbial iron access.

Introduction

The natural resistance-associated macrophage protein 1, Slc11a1 (solute carrier family 11 member 1, formerly Nramp1), modifies macrophage activation, thus conferring innate resistance to intracellular pathogens as well as susceptibility to certain autoimmune diseases (Blackwell et al., 1999, Canonne-Hergaux et al., 1999). Slc11a1 is an integral membrane protein expressed in the late phagolysosomes of macrophages and in intracellular compartments of neutrophils and dendritic cells (Gruenheid et al., 1999, Valdez et al., 2008a). Although Slc11a1 primarily acts as a proton-dependent transporter for divalent ions such as iron or manganese, it also exerts pleiotropic effects on innate immune functions. For instance, Slc11a1 modulates the production of chemokines and cytokines such as macrophage inflammatory protein 1-alpha (Mip-1α), tumour necrosis factor-alpha (TNF-α), interleukin (IL)-1β, and the formation of reactive oxygen and nitrogen species (ROS and RNS, respectively), as well as antigen processing and presentation (Blackwell et al., 2003, Forbes et al., 2001, Kuhn et al., 1999, Nevo et al., 2006, Fritsche et al., 2003, Valdez et al., 2008b). Slc11a1-mediated resistance to intramacrophage microbes covers distinct taxonomically unrelated pathogens such as Salmonella, Mycobacterium and Leishmania species, and part of this appears to be linked to its capacity to transport iron (Forbes et al., 2001, Huynh et al., 2008). While Slc11a1 may shift iron into phagolysosomes (Zwilling et al., 1999, Kuhn et al., 1999, Kuhn et al., 2001), where redox-active ferrous iron may directly support the formation of ROS in the Fenton reaction, evidence is abundant that Slc11a1 expression is associated with iron efflux out of phagolysosomes and subsequently out of macrophages (Atkinson et al., 1999, Mulero et al., 2002, Fritsche et al., 2007), supporting the hypothesis that Slc11a1 may deprive intracellular microorganisms of essential iron.

Notably, the efficacy of several macrophage immune functions is strongly influenced by iron. Whereas trace amounts of iron are required for the generation of ROS, several pathways of host protection supported by mononuclear phagocytes, including those driven by interferon-gamma (IFN-γ), are impaired by elevated macrophage iron content (Weiss, 2002). For instance, macrophage iron overload results in reduced expression of TNF-α, inducible nitric oxide synthase (iNOS) and major histocompatibility complex (MHC) class II, thus simultaneously impairing cytokine production, bacterial elimination via RNS and antigen presentation (Weiss et al., 1994, Oexle et al., 2003, Nairz et al., 2007, Recalcati et al., 1998). Importantly, TNF-α signalling and iNOS expression are of pivotal importance in innate immunity against Salmonella Typhimurium (Vazquez-Torres et al., 2000, Vazquez-Torres et al., 2001, Mastroeni et al., 2000). Since iron exerts inhibitory effects on these antimicrobial pathways, iron overload of the mononuclear phagocyte system is associated with an increased incidence of infections, as seen in secondary iron overload (Nairz et al., 2006, Rahav et al., 2006, Wang et al., 2003).

Given the importance of iron for innate immune functions and metabolism, macrophages tightly control cellular iron homeostasis by regulating both transmembraneous iron fluxes and intracellular iron storage (Knutson et al., 2003a, Weiss, 2005, Theurl et al., 2005, Oppenheimer, 2001). Phagocytes can acquire non-transferrin-bound iron (NTBI) via divalent metal transporter 1 (Dmt1) and transferrin-bound iron (TBI) via transferrin receptor 1 (TfR1). The phagocytosis of senescent erythrocytes provides macrophages with heme-iron, which is released upon enzymatic degradation by heme oxygenase 1 (Hmox1). Surplus iron can be either stored within ferritin or exported out of macrophages via ferroportin 1 (Fpn1), thus being recycled to the circulating pool of TBI (Bonnah et al., 2004, Nairz et al., 2006, Hentze et al., 2004).

As a consequence of its involvement in central metabolic processes, iron directly promotes the growth and proliferation of microorganisms by serving as an essential nutrient (Weinberg, 1998, Weinberg, 2000, Schaible et al., 2005). In fact, with the sole exception of Borrelia burgdorferi, which uses manganese in place of iron-sulfur clusters in its metabolic enzymes, all mammalian pathogens require iron (Schaible et al., 2004). Accordingly, microbial iron uptake mechanisms promote virulence whereas genetic deletion of iron import pathways reduces pathogenicity (Boyer et al., 2002, Janakiraman et al., 2000, Fischbach et al., 2006, Crouch et al., 2008, Jung et al., 2008, Schrettl et al., 2007, Winkelmann, 2002). The limitation of iron availability for pathogens is thus an effective strategy of innate host defense as shown in several in vitro and in vivo studies (Ibrahim et al., 2007, Nairz et al., 2007, Olakanmi et al., 2000, Paradkar et al., 2008, Kanti Das et al., 2008).

To better understand the pleiotropic functions of Slc11a1 in host protection against intracellular pathogens, we have initiated in vitro experiments to simultaneously study iron homeostasis and effector functions in RAW264.7 macrophage-like cells stably transfected with functional (RAW-37 cells) or non-functional (RAW-21 cells) Slc11a1 during infection with S. Typhimurium.

Results

Effects of Slc11a1 on iron acquisition, iron mobilization and iron release by Salmonella-infected macrophages

When investigating macrophage iron import, we found that TfR1 mRNA levels and transferrin-mediated iron uptake were significantly increased in RAW-21 cells lacking functional Slc11a1 when compared with Slc11a1-expressing RAW-37 macrophages. These changes were also seen in Salmonella-infected and/or IFN-γ treated cells (Figures 1A and B). In contrast, neither the expression of Dmt1 mRNA nor the uptake of NTBI differed significantly between Slc11a1-non-expressing RAW-21 and Slc11a1-expressing RAW-37 cells (Supplementary figures S1A and B). Moreover, while Fpn1 mRNA levels were not different between solvent-treated RAW-21 and RAW-37 cells, Salmonella infection resulted in approximately 80% lower Fpn1 mRNA expression in Slc11a1-dysfunctional RAW-21 cells (P < 0.05). This difference was also apparent in Salmonella-infected phagocytes subsequently stimulated with IFN-γ (Figure 1C). In parallel with the reduced Fpn1 expression, cellular iron efflux was approximately 95–100% lower in Salmonella-infected RAW-21 cells with or without IFN-γ-stimulation, in comparison to Slc11a1-expressing RAW-37 cells subjected to the same treatment (Figure 1D). The observed changes in TfR1 and Fpn1 mRNA levels were similar to their levels of protein expression (Figure 2). Of relevance, levels of hepcidin antimicrobial peptide (Hamp) mRNA were not altered upon S. Typhimurium infection and did not differ between RAW-21 and RAW-37 cells (data not shown).

Figure 1. Effects of Slc11a1 on iron acquisition and iron release by Salmonella-infected macrophages.

Slc11a1-non-functional RAW-21 cells (light grey bars) and Slc11a1-functional RAW-37 cells (dark grey bars) were infected with S. Typhimurium (S. tm.) at a MOI of 10 and stimulated with 50 U/ml IFN-γ for 24 hours. TfR1 and Fpn1 mRNA levels (Figures 1A and 1C, respectively) were determined by qPCR. Values were corrected for the amount of 18S ribosomal RNA, which was determined in parallel. Results are shown as relative differences of this ratio in comparison to unstimulated Slc11a1-expressing RAW-37 control macrophages (=1.0). Data are expressed as mean ± SD of five independent experiments. The uptake of TBI as well as iron release (depicted in Figures 1B and 1D and, respectively) were determined as described in ‘Experimental procedures’. Data are shown as mean ± SD of five independent experiments performed in duplicates and are expressed as fold change in the relative iron uptake/release in comparison to RAW-37 controls (=1.0). * P < 0.05 for infected/stimulated RAW-21 macrophages in comparison to RAW-21 controls; ° P < 0.05 for infected/stimulated RAW-37 in comparison to untreated RAW-37 cells; # P < 0.05 for the comparison of RAW-37 and RAW-21 cells subjected to the same treatment.

Figure 2. Regulation of TfR1 and Fpn1 protein levels in Salmonella-infected RAW-21 and RAW-37 cells.

Slc11a1-non-expressing RAW-21 cells and Slc11a1-expressing RAW-37 cells were infected with Salmonella and stimulated with IFN-γ exactly as described in the legend to Figure 1. Whole cell lysates were analysed by immunoblotting using specific antibodies to TfR1 (upper panel) and Fpn1 (middle panel). Equal loading of protein extracts was confirmed by reprobing membranes with an anti-Actin antibody (lower panel). One of four representative immunoblot experiments is shown.

After 24 hours, Hmox1 transcripts were significantly induced in Salmonella-infected phagocytes and further enhanced upon addition of IFN-γ. However, Hmox1 mRNA and protein levels were comparable in RAW-21 and RAW-37 cells. Similarly, mRNA levels of the heme-exporting feline leukaemia virus, subgroup C, receptor (Flvcr) were not different between the two cell lines (data not shown).

Effects of Slc11a1 on cellular iron content and on intracellular iron traffic in Salmonella-infected macrophages

To examine the effects of Slc11a1 on iron availability within Salmonella-infected macrophages, we first examined the expression of the iron storage protein ferritin. Salmonella infection resulted in a marked enhancement of ferritin mRNA and protein expression in both, RAW-21 and RAW-37 cells, in comparison to unstimulated cells. However, Slc11a1 did not influence ferritin expression as no difference was observed between these two cell lines under any of the experimental conditions (Figures 3A and B).

Figure 3. Effects of Slc11a1 on ferritin expression and on intracellular iron content in Salmonella-infected macrophages.

RAW-21 and RAW-37 cells were infected with Salmonella and activated with IFN-γ exactly as described in the legend to Figure 1. The mRNA levels of H-ferritin were determined by qPCR (Figure 3A). Results are shown as relative differences of this ratio in comparison to the unstimulated control macrophages (=1.0). The total cellular iron content was determined by means of atomic absorption spectrometry and normalized for the protein content (Figure 3B). Data are expressed as means ± SD of five independent experiments. Statistically significant differences are indicated as described in the legend to figure 1.

Nevertheless, as ferritin expression at the protein level under inflammatory conditions is influenced by both iron concentrations and inflammatory stimuli, we determined cellular iron content by means of atomic absorption spectrometry. Iron content decreased following Salmonella infection in both cells types, but RAW-21 cells were found to have a 43–63% higher iron content than Slc11a1-expressing RAW-37 cells under the same conditions (P < 0.05, Figure 3C).

Impact of Slc11a1 on macrophage immune effector functions

To better understand the effects of Slc11a1 on innate immune effector functions, we determined the concentrations of macrophage-derived mediators in culture supernatants of RAW-21 cells and Slc11a1-expressing RAW-37 stimulated with IFN-γ and/or infected with Salmonella. Notably, the production of RNS, measured as the oxidation end-product nitrite in cell culture supernatants, was significantly lower in RAW-21 cells than in Slc11a1-expresssing RAW-37 cells (Figure 4A). This may be due to a transcriptional regulation as we found significantly decreased expression of inducible nitric oxide synthase (iNOS) mRNA in Salmonella-infected RAW-21 macrophages in comparison to RAW-37 cells (Supplementary figure S2A). In contrast, Slc11a1 had no effect on the transcription of phox-p47 nor the production of ROS, the latter determined by measurement of total oxidative capacity (TOC) in cell culture supernatants (data not shown and supplementary figure S2B). Intriguingly, the presence of Slc11a1 also correlated with the expression of several macrophage-derived cytokines. Upon infection with Salmonella and with or without the addition of IFN-γ, RAW-21 cells lacking functional Slc11a1 secreted significantly lower amounts of TNF-α and IL-6 (Figures 4B and 4C) and substantially higher amounts of IL-10 in comparison to Slc11a1-expressing RAW-37 macrophages treated identically (Figure 4D). In contrast, IL-1β levels in cell culture supernatants did not differ markedly between the two cell lines (Supplementary figure S2C). Similarly, Slc11a1 did not influence levels of IL-18 or IL-23 mRNA (data not shown). Of interest, RAW-21 also displayed lower mRNA levels of chemokines monocyte chemoattractant protein (Mcp)-1 and Mip-1α, whereas no effect of Slc11a1 was observed with respect to Ccl-5 or Cxcl-10 mRNA expression (data not shown).

Figure 4. Impact of Slc11a1 on macrophage immune effector functions after IFN-γ stimulation.

RAW-21 and RAW-37 cells were stimulated with IFN-γ and/or infected with Salmonella for 24 hours. Culture supernatants were analyzed for RNS (as determined by measuring nitrite levels, Figure 4A), TNF-α (Figure 4B), IL-6 (Figure 4C) and IL-10 (Figure 4D) as described in ‘Experimental procedures’. Data are shown as mean ± SD of four to five independent experiments. Statistically significant differences are indicated as described in the legend to figure 1.

Electrophoretic mobility shift assays were performed to determine whether the impaired production of pro-inflammatory mediators by Salmonella-infected and IFN-γ stimulated RAW-21 cells could be attributed to specific transcriptional regulators. Slc11a1-non-expressing RAW-21 cells exhibited higher binding affinities of the transcription factor signal transducer and activator of transcription (STAT)-3 in response to Salmonella infection and/or cytokine stimulation for 24 hours (Supplementary figure S3), as recently reported for IFN-γ/LPS stimulation (Fritsche et al., 2008). In contrast, activation of interferon regulatory factor (IRF)-1, nuclear factor (NF)-κB and STAT-1 was not affected by the presence of functional Slc11a1 (Supplementary figure S3 and data not shown).

Impact of Slc11a1 on macrophage antimicrobial activity upon IFN-γ stimulation and iron perturbations

We observed increased bacterial survival in Slc11a1-non-expressing RAW-21 cells which became significant by 16 to 24 hours post-infection (Figure 5A), in agreement with previous observations (Blackwell et al., 2001, Forbes et al., 2001, Fritsche et al., 2008). Stimulation with IFN-γ significantly reduced Salmonella survival in both cell types (Figure 5B). TBI supplementation resulted in a dose-dependent increase in Salmonella survival within RAW-21 mononuclear phagocytes, while the presence of Slc11a1 in RAW-37 cells efficiently abrogated this Salmonella growth-promoting effect of iron supplementation (Figure 6A). In parallel, TBI dose-dependently reduced the generation of nitrite as well as the secretion of TNF-α by Salmonella-infected RAW-21 macrophages, which was not observed in Slc11a1-expressing RAW-37 phagocytes (Figures 6B and 6C). Interestingly, TBI supplementation dose-dependently augmented IL-10 expression in infected RAW-21 cells, but the low IL-10 levels seen in Slc11a1-functional RAW-37 cells were not changed by iron supplementation (Figure 6D). NTBI, delivered as ferrous sulfate, exerted similar effects on intracellular Salmonella survival and macrophage nitrite and cytokine production to supplementation with TBI (data not shown).

Figure 5. Impact of Slc11a1 on the intracellular survival of Salmonella.

RAW-21 and RAW-37 cells were infected with Salmonella for the indicated periods (Figure 5A) or infected for 24 hours and activated with IFN-γ (Figure 5B). Bacterial load was determined by selective plating as described in ‘Experimental procedures’.

Figure 6. Impact of Slc11a1 on macrophage immune effector functions in the presence of exogenous iron perturbations.

In additional experiments, macrophages were infected with Salmonella and exposed to serial concentrations of human holo-transferrin. Bacterial load was determined by selective plating (Figure 6A). Cell culture supernatants were analyzed for the formation of nitrite, TNF-α, IL-10 and IL-6 as described (Figures 6B, C, D and E, respectively). Statistically significant differences are indicated as described

Effects of synthetic Hamp and of an anti-TfR1 antibody on Salmonella-infected RAW-21 and RAW-37 macrophages

We next investigated the effects of synthetic Hamp and of a blocking anti-TfR1 antibody, both of which are expected to modulate macrophage transmembraneous iron fluxes, on effector functions of RAW-21 cells as compared to Slc11a1-functional RAW-37 phagocytes. Interestingly, the addition of Hamp to Salmonella-infected macrophages increased the bacterial load by 55–60% in both RAW-21 and RAW-37 cells, in comparison to cells not treated with Hamp (Figure 7A). However, antibody-mediated blockade of TfR1 reduced the number of intracellular bacteria only in RAW-21, but not in Slc11a1-expressing RAW-37 macrophages. Notably, increased concentrations of nitrite were measured in both RAW-21 and RAW-37 macrophages following the addition of anti-TfR1 antibody (Figure 7B). In contrast, the addition of ferrous sulfate and Hamp impaired the generation of nitrite exclusively in Salmonella-infected RAW-21 cells, but not in infected Slc11a1-competent RAW-37 cells. Of note, Hamp resulted in significant inhibition of TNF-α production in both macrophage cell lines, whereas the addition of a blocking anti-TfR1 antibody had no substantial effect on the production of TNF–α in either RAW-21 or RAW-37 cells (Figure 7C).

Figure 7. Contributions of Fpn1-mediated iron export and of TBI-uptake to Slc11a1-dependent macrophage functions.

RAW-21 and RAW-37 cells were pre-treated for 8 hours with FeSO₄, DFO, synthetic Hamp, a blocking α-TfR1 antibody or the appropriate control. Subsequently, cells were infected with Salmonella for 12 hours. Intracellular bacteria were enumerated (Figure 7A) and concentrations of TNF-α (Figure 7B) and of nitrite (Figure 7C) in supernatants were measured by a specific ELISA or the Griess reaction, respectively.

Limitation of iron availability for intracellular Salmonella in the presence of Slc11a1

To see whether the alterations of transmembrane iron fluxes mediated by functional Slc11a1 may affect iron acquisition by intracellular bacteria, we directly measured the uptake of ⁵⁹Fe, delivered as either NTBI or TBI, into internalized Salmonella. Notably, the incorporation of NTBI and TBI by intramacrophage bacteria was significantly higher in RAW-21 cells in comparison to Salmonella residing within RAW-37 cells expressing functional Slc11a1 (Figures 8A and 8B). Of note, IFN-γ was able to further reduce the availability of both iron forms to phagocytosed Salmonella independently of Slc11a1.

Figure 8. Limitation of iron availability for intramacrophage Salmonella in the presence of Slc11a1.

The acquisition of NTBI (Figure 8A) and TBI (Figure 8B) by intramacrophage bacilli was determined as described in ‘Experimental procedures’ using S. Typhimurium engulfed by Slc11a1-non-functional RAW-21 in comparison to Slc11a1-functional RAW-37 phagocytes. Data are shown as means ± SD of five independent experiments and normalized for the number of bacteria as determined by selective plating.

Effects of bacterial iron acquisition systems on effector functions of RAW-21 and RAW-37 macrophages

The effect of Slc11a1 on the survival of S. Typhimurium strains carrying mutations in entC, required for siderophore biosynthesis, with or without mutations in the sit and feo loci, required for siderophore-independent iron acquisition, was determined. Wild-type S. Typhimurium exhibited greater intracellular survival than mutant strains lacking siderophore biosynthesis, and this difference was most pronounced in RAW-21 cells (Figure 9A). RAW-21 cells controlled the replication and survival of all S. Typhimurium strains less efficiently than did Slc11a1-expressing RAW-37 cells, although significant differences were only observed upon infection with wild-type Salmonella (Figure 9A). Interestingly, the intracellular growth of both wild-type and isogenic entC mutant Salmonella strains was enhanced following addition of holo-transferrin to infected Slc11a1-dysfunctional RAW-21 macrophages, whereas the entC sit feo triple mutant showed no effect. In contrast, RAW-37 cells limited the growth of all bacterial strains whether or not supplemental holo-transferrin was provided.

Figure 9. Antibacterial effects in RAW-21 and RAW-37 cells infected with iron-uptake mutant Salmonella.

RAW-21 and RAW-37 phagocytes were infected with S. Typhimurium wild-type strain as well as with isogenic entC and entC sit feo mutant derivatives and treated with PBS or 12.5 μg/ml holo-transferrin. The intracellular survival of bacteria was enumerated after 20 hours, and data are presented as relative values in comparison to solvent-treated RAW-37 cells infected with wild-type Salmonella (Figure 9A). Culture supernatants were analyzed for the concentrations of TNF-α (Figure 9B) and nitrite (9C).

The production of nitrite and IL-10 was not affected by entC, sit or feo mutations (Figure 9B and supplementary figure S4), but higher TNF-α levels were elicited by entC and entC sit feo mutants following infection of RAW-21 cells (Figure 9C). Notably, holo-transferrin inhibited nitrite and TNF-α production by infected Slc11a1-deficient RAW-21 phagocytes but did not appear to affect the lower levels produced by RAW-37 cells (Figures 9B and C).

Discussion

Herein, we provide evidence that Slc11a1 exerts multiple functions within Salmonella-infected macrophages, simultaneously influencing both, the expression of antimicrobial immune responses and the access of intracellular bacteria to cellular iron. Specifically, the expression of Slc11a1 reduces intracellular iron levels of Salmonella-infected macrophages, which is exerted by at least two pathways: first, a reduction in TfR1-mediated iron uptake and second, the induction of Fpn1-mediated iron efflux (Figure 1). These events limit the availability of iron for Salmonella residing within phagosomal vacuoles. In addition, due to regulatory effects of iron on IFN-γ-mediated pathways in macrophages, the reduction in intracellular iron concentration promotes the production of TNF-α and IL-6 as well as the expression of iNOS (Oexle et al., 2003, Weiss et al., 1994, Nairz et al., 2008). In parallel, Slc11a1 either directly or via modulation of pro-inflammatory immune effector pathways suppresses the secretion of IL-10 (Fritsche et al., 2008). Accordingly, Slc11a1-mediated resistance against leishmanial parasites, mycobacteria and salmonellae is associated with a T helper cell type 1 (Th1) response, which favours host defense against intracellular microorganisms (Soo et al., 1998, Kramnik et al., 1994, Valdez et al., 2008a, Valdez et al., 2008b). It is tempting to speculate that the increased production of pro-inflammatory cytokines in conjunction with the reduced production of the anti-inflammatory mediator IL-10 may contribute to both the increased resistance to infection and the increased susceptibility to autoimmune disorders that are associated with Slc11a1 in both mice and humans (Blackwell et al., 2003).

Intriguingly, there were differences in the iron-dependent influence of Slc11a1 on the generation of nitrite, TNF-α, IL-6 and IL-10 in Salmonella-infected macrophages (Figure 6)., While the addition of TBI dose-dependently inhibited the generation of nitrite and TNF-α exclusively in RAW-21 phagocytes lacking Slc11a1, Slc11a1-expressing RAW-37 cells were capable of producing high levels of RNS and TNF-α and of secreting even larger amounts of IL-6 upon gradually increasing TBI supplementation. The opposite was true with IL-10 formation. While Slc11a1-positive cells could limit IL-10 secretion in the face of increasing iron levels, Salmonella-infected Slc11a1-negative macrophages responded to the exposition to high concentrations of TBI by producing increasing amounts of IL-10. These differential effects were also observed in response to stimulation with IFN-γ and/or LPS (Fritsche et al., 2008), suggesting that they may not indirectly result from the increased Salmonella burden within RAW-21 cells. Being aware of the differences in iron handling between Slc11a1-positive and -negative macrophages and of the fact that iron differently affects pro- and anti-inflammatory immune pathways, these observations can thus be traced back to limitation of labile cytoplasmic iron by Slc11a1-expressing cells. Accordingly, we also suggest that Slc11a1-expressing cells can rapidly shift iron from the cytoplasm out of cells via Fpn1. In contrast, iron accumulates in Slc11a1-negative cells and then exerts its negative effects on pro-inflammatory immune effectors pathways while promoting the formation of IL-10 (Fritsche et al., 2008, Knutson et al., 2003b, Mencacci et al., 1997, Oexle et al., 2003, Oppenheimer, 2001, Weiss et al., 1999).

To date, the underlying molecular mechanisms by which Slc11a1 concomitantly increases the production of TNF-α, IL-6 and RNS and differently affects the expression of immune genes have not been elucidated.. Since TNF-α, IL-6, Mcp-1, Mip-1α and iNOS mRNA levels are substantially higher in Slc11a1-expressing RAW-37 cells (Supplementary figure S2 and additional data not shown), a transcriptional or post-transcriptional mechanism may be involved. Notably, Salmonella-infected and/or IFN-γ-stimulated RAW-21 and RAW-37 macrophages showed comparable binding activities of the transcription factors IRF-1, NF-κB and STAT-1, all of which are pivotal for the up-regulation of iNOS in response to inflammatory stimuli (Fritsche et al., 2003). However, these differences were only observed at late time-points, whereas differential cytokine mRNA expression was evident earlier. Therefore, the increased activation of STAT-3 in LPS- or Salmonella-stimulated RAW-21 macrophages is presumably a consequence rather than a cause of the higher IL-10 levels produced by these cells. In this context, it is of considerable interest that the neutralization of IL-10 by specific antibodies restored the induction of TNF-α and iNOS and enhanced Salmonella killing of Salmonella in Slc11a1-deficient RAW-21 cells (Fritsche et al., 2008). Some investigators have suggested that Slc11a1 may promote TNF-α and RNS formation by prolonging the half-life of TNF-α and iNOS mRNAs, respectively (Brown et al., 1997), although we were unable to corroborate this observation (Fritsche et al., 2003). Moreover, it is possible that TNF-α produced in greater quantities by Slc11a1-positive macrophages contributes to their higher capacity to generate RNS by acting as a co-activator of iNOS expression (Ables et al., 2001). Thus, Slc11a1 may regulate cytokine expression via activation of iron- or NO-sensitive transcription factors, although we did not detect differential activations of NF-κB and NF-IL6 (Supplementary figure S3 and details not shown). Alternatively, it is possible that functional Slc11a1 may selectively stimulate certain macrophage functions by its ability to promote the maturation of Salmonella-containing phagosomes (Cuellar-Mata et al., 2002, Jabado et al., 2003). Intracellular fusion events involving pathogen-containing vesicles may initiate signal transduction mechanisms ensuring a predominately pro-inflammatory response in Slc11a1-expressing infected macrophages. Anyhow, regulatory interactions between iron availability and cellular immune effector pathways are likely to account for some of the effects of Slc11a1 on host resistance to diverse intracellular pathogens such as Salmonella, Mycobacterium or Leishmania spp. (Diefenbach et al., 1999, Mastroeni et al., 2000, Weiss et al., 1994)

Since phagocytes are a preferred habitat for S. Typhimurium (Vazquez-Torres et al., 1999), the control of iron homeostasis in infected mononuclear cells is crucial for effective host defense against this bacterium (Collins, 2003, Theurl et al., 2005). Within macrophages, Slc11a1 is found exclusively in the membrane of late endosomes and phagolysosomes but has not been detected in the plasma membrane (Blackwell et al., 2003, Searle et al., 1998, Atkinson et al., 1997, Gruenheid et al., 1999, Gruenheid et al., 1997, Canonne-Hergaux et al., 2002, Lam-Yuk-Tseung et al., 2006). Therefore, the effect of Slc11a1 on cellular iron release is presumably an indirect one. This may be a consequence of the differences in immune activation between Scl11a1-positive and -negative macrophages, since increased formation of NO and oxygen radicals, as seen in Slc11a1-expressing phagocytes, can directly affect the orchestration of cellular iron homeostasis via iron regulatory proteins (IRPs). Specifically, NO and oxygen radicals modify IRP binding affinity to IREs thereby affecting ferritin and TfR1 mRNA processing (Muckenthaler et al., 2008, Recalcati et al., 1998, Weiss et al., 1993). Our results show reduced acquisition of iron by intraphagolysosomal Salmonella residing within Slc11a1-functional cells, which strongly supports the concept that Slc11a1 shifts iron from the phagolysosomal compartment to the cytosolic space. An intermittent cytosolic accumulation of iron may affect Fpn1 and TfR1 expression via both transcriptional and IRP/IRE-mediated posttranscriptional control (Figures 1 and 2). In addition, the presence of Fpn1 in the cell surface membrane can be regulated at the posttranslational level by Hamp produced by activated macrophages in an autocrine fashion (Theurl et al., 2008). However, the cytosolic space does not appear to be the final resting location for this iron, as in steady state conditions, a fall in TfR1-mediated iron uptake along with increased Fpn1-mediated iron export markedly reduced total macrophage iron content and the cytoplasmic labile iron pool, as measured by atomic absorption spectrometry (Figure 4) and electrophoretic mobility shift assays of IRP binding (Atkinson et al., 1999, Barton et al., 1999, Fritsche et al., 2007, Nairz et al., unpublished observations). Slc11a1 did not affect the expression of ferritin or the incorporation of ⁵⁹Fe into ferritin multimers (Figure 3 and data not shown). Furthermore, we did not detect any effect of Slc11a1 on the expression of the heme-mobilizing enzyme Hmox1 or the heme-exporter Flvcr in infected macrophages. Although we did not challenge macrophages with heme-iron, hemoglobin or erythrocytes, our results support the idea that Slc11a1 modulates the cytosolic iron content. Iron-sensitive fluorogenic probes targeting specific organelles may be helpful in investigating Slc11a1-mediated iron fluxes between intracellular compartments under real-time conditions by means of microfluorescence imaging (Glickstein et al., 2005, Jabado et al., 2000). However, Slc11a1 leads to reduction of cytoplasmic iron levels and withholds the metal from pathogens residing within phagolysosomes. As Hamp mRNA levels were not affected by the presence or absence of Slc11a1, it is highly unlikely that changes in Hamp expression are responsible for the differences in iron export observed between RAW-21 and Slc11a1-expressing RAW-37 macrophages. Although the blockade of Fpn1-mediated iron efflux by the addition of synthetic Hamp resulted in a marked increase in the microbial burden of infected phagocytes, Hamp treatment of infected macrophages failed to abolish the differences in bacterial survival between RAW-21 and RAW-37 cells, suggesting that Slc11a1-dependent antimicrobial actions are not dependent on Fpn1-mediated iron export (Figure 7).

A number of studies have attempted to elucidate the biochemical function of Slc11a1, yet its mode of action remains a topic of some debate. Experiments with isolated phagosomes suggested that Slc11a1 might transport iron into these organelles to provide a catalyst for the production of cytotoxic ROS via the Fenton reaction (Kuhn et al., 1999, Kuhn et al., 2001, Zwilling et al., 1999). However, investigations using intact cells rather indicated that Slc11a1 fluxes divalent ions such as iron and manganese out of the lumen of phagosomes into the cytosol (Jabado et al., 2002, Gomes et al., 1998, Forbes et al., 2003). Moreover, the topology of Slc11a1 in the vacuolar membrane relative to the proton gradient indicates that the direction of cation transport is outward. Although we did not directly measure iron transport across the phagolysosomal membrane or iron concentrations within the phagolysosome, our demonstration of reduced iron acquisition by intraphagolysosomal Salmonella residing within Slc11a1-expressing cells is most consistent with the idea that Slc11a1 depletes the phagolysosome of iron. To the best of our knowledge, this is the first report to directly compare ⁵⁹Fe acquisition by an intraphagolysosomal pathogen in the presence or absence of functional Slc11a1. Studies with Slc11a1^Gly169 (wildtype) and Slc11a1^Asp169 (mutant/functional null) mice and with peritoneal macrophages isolated from these will be of interest to see whether the effect of Slc11a1 is also relevant in primary macrophages and the in vivo situation. This apprears likely since a recent study employing Slc11a1/Nramp11^+/+ and Slc11a1/Nramp1^−/− mice in a model of Salmonella-induced colitis demonstrated that the Slc11a1 phenotype was associated with a pro-inflammatory immune response and an increased clearance of the bacteria (Valdez et al., 2008b).

Of note, it has been demonstrated that Leishmania amazonensis is capable of up-regulating its ferrous iron importer LIT1 within phagolysosomes and that this regulation is accelerated in response to iron starvation by functional Slc11a1 (Huynh et al., 2006). Whether or not this also applies to intraphagolysosomal Salmonella is currently unknown. However, Salmonella expresses a set of virulence genes encoded within Salmonella pathogenicity island 2 as a specific response to Slc11a1-mediated iron starvation (Zaharik et al., 2002). In addition, Mycobacterium tuberculosis reportedly enhances siderophore production upon Slc11a1-mediated iron deprivation (Jabado et al., 2003).

The divalent cation transporters SitABCD and MntH have been shown to promote Salmonella growth in macrophages expressing Slc11a1 (Zaharik et al., 2004). Here we show for the first time that EntC-dependent catecholate-type siderophore production promotes Salmonella growth in both Slc11a1-expressing and Slc11a1-deficient macrophages (Figure 8A). The ability of RAW-37 cells to effectively restrict wild type S. Typhimurium growth in phagocytes illustrates the capacity of Slc11a1 to limit the availability of cations in the vacuolar compartment beyond even the ability of high-affinity microbial iron acquisition systems to obtain them.

Since iron plays a central role in the interaction between host and pathogen, affecting both microbial replication and the activation of cell-mediated immune pathways, a better understanding of Slc11a1 transporter function and the further elucidation of the mechanisms linking metal ion redistribution to pro- and anti-inflammatory gene expression might enable the development of pharmacological strategies to specifically target Slc11a1 as a novel approach to control infections with intracellular pathogens such as Salmonella and Mycobacterium tuberculosis.

We have provided evidence to support a dual role of the phagolysosomal transport protein Slc11a1 in promoting host resistance against intracellular microbes such as S. Typhimurium. By reducing the intracellular iron content of infected macrophages, Slc11a1 starves ingested bacteria of iron while concomitantly promoting the generation of antimicrobial effector molecules. These properties of Slc11a1 provide important insights into mechanisms by which nutrient withholding contributes to innate immunity.

Experimental procedures

Cell culture

RAW264.7 murine macrophage-like cells were originally isolated from Slc11a1-nonfunctional BALB/c mice and obtained from the American Type Culture Collection. RAW264.7 cells were stably transfected by standard methods using the pHβA-1-neo expression plasmid containing the full-length Slc11a1 cDNA (RAW-37) or an antisense-Slc11a1 construct (RAW-21) as described previously (Atkinson et al., 1999). Routinely, cells were grown in DMEM (containing 0.1 mg/ml Fe(NO₃)₃ · 9H₂O; obtained from Biochrom, Berlin, Germany) containing 10% FCS, 2 mM L-glutamine and 50 μg/ml ampicillin at 37°C in humidified air containing 5% CO₂ and periodically selected for plasmid retention using G418 (Invitrogen). Stable expression of Slc11a1 in RAW-37 phagocytes was confirmed by immunoblotting (Supplementary figure S5, upper panel).

For infection experiments, 1×10⁶ RAW-21/37 cells were seeded into 6-well plates in 2 ml of DMEM containing 10% FCS, 2 mM L-glutamine and 50 μg/ml ampicillin. S. Typhimurium strain C5RP4 (kindly provided by Dr. Pietro Mastroeni, Cambridge University, Cambridge, United Kingdom), a virulent wild-type isolate carrying natural resistance to ampicillin, was used for experiments unless otherwise indicated and grown in Luria-Bertani (LB) broth (obtained from Sigma) containing 50 μg/ml ampicillin to late-logarithmic phase. After preincubation of S. Typhimurium in complete DMEM at 37°C for 20 minutes, RAW-21/37 cells were infected with S. Typhimurium at a multiplicity of infection (MOI) of 10 for 1 hour exactly as described (Nairz et al., 2007). Control cells were treated with complete DMEM without any microbes. After 1 hour, RAW-21/37 cells were washed three times with phosphate-buffered saline (PBS; purchased from Invitrogen) and incubated in 2 ml of complete DMEM additionally containing 20 μg/ml of gentamicin (Gibco) in order to kill extracellular bacteria. Thereafter, macrophages were stimulated with recombinant murine IFN-γ (50 U/ml; obtained from R&D), whereas controls were treated with PBS as solvent. After incubation for an additional 23 hours, RAW-21/37 cells were washed three times in PBS and subjected to RNA or protein isolation. Alternatively, cells were treated with lipopolysaccharide (LPS; obtained from Sigma) and IFN-γ and subjected to RNA isolation.

For certain experiments, RAW-21/37 cells were cultured in complete DMEM without any antibiotics and infected with S. Typhimurium strain MLC446 (wild-type; corresponding to ATCC 14028s) or isogenic mutant derivatives MLC619 (entC::aph; single mutant) and MLC774 (entC::aph sit::bla feo::Tn10 (Tet^r); triple mutant). Mutant strains were constructed and grown as described (Crouch et al., 2008) and used as detailed above.

For quantification of intracellular bacteria by means of gentamicin protection, RAW-21/37 cells were infected as described above and stimulated with IFN-γ, serial concentrations of human holo-transferrin or ferrous sulfate (all from Sigma). In experiments using synthetic murine hepcidin 1 (1 μM; purchased from PeptaNova), a blocking monoclonal anti-TfR1 antibody and desferrioxamine (DFO; added at a concentration of 50 μM and purchased from Sigma), cells were pre-treated for 8 hours and subsequently infected for an additional 12 hours. The appropriate solvent or isotype control antibody served as controls. After an infection period of 24 hours, cells were washed 5 times with PBS, lysed in 0.5% deoxycholic acid (Sigma) and plated under sterile conditions in appropriate dilutions onto LB agar plates.

Quantitative Reverse Transcription PCR (qPCR)

Preparation of total RNA from RAW-21/37 macrophages was performed by a guanidinium-isothiocyanate-phenol-chloroform based extraction method exactly as described (Ludwiczek et al., 2007). Quantification of mRNA expression was carried out by means of quantitative reverse transcription PCR (qPCR), which was performed exactly as described (Ludwiczek et al., 2004). Amplifications conditions were as follows: 95°C for 10 min, 40 cycles of 95°C for 15 s, 60°C for 60 s. The following primers and TaqMan probes (Microsynth), the latter carrying 5′-FAM and 3′-BHQ1-labels, were used: mu-Dmt1: 5′-CCAGCCAGTAAGTTCAAGGATCC-3′, 5′-GCGTAGCAGCTGATCTGGG-3′, 5′-TGGCCTCGCGCCCCAACA-3′, mu-TfR1: 5′-CGCTTTGGGTGCTGGTG-3′, 5-GGGCAAGTTTCAACAGAAGACC-3′, 5′-CCCACACTGGACTTCGCCGCA-3′, mu-Fpn1: 5′-CTACCATTAGAAGGATTGACCAGCT-3 ′, 5 ′-CAAATGTCATAATCTGGCCGA-3′, 5′-CAACATCCTGGCCCCCATGGC-3′, mu-Hamp: 5′-TGTCTCCTGCTTCTCCTCCTTG-3 ′, 5 ′-AGCTCTGTAGTCTGTCTCATCTGTTG-3′, 5′-CAGCCTGAGCAGCACCACCTATCTCC-3′, mu-Hmox1: 5′-GTGATGGAGCGTCCACAGC-3′, 5′-TGGTGGCCTCCTTCAAGG-3′, 5′-CGACAGCATGCCCCAGGATTTGTC-3′, mu-H-ferritin: 5′-GCGAGGTGGCCGAATCT-3′, 5′-CAGCCCGCTCTCCCAGT-3′, 5′-CCTGCAGGATATAAAGAAACCAGACCGTGA-3′, mu-Flvcr: 5 ′-ATCTGGAACCTGTGCAGAAACA-3 ′, 5 ′-ATTGAATAAAATGCTCCAGTCATGAT-3′, 5′-CCCCTTTGTTCTCCTGCTGGTCAGTTATG-3′, mu-iNOS: 5′-CAGCTGGGCTGTACAAACCTT-3′, 5′-CATTGGAAGTGAAGCGTTTCG-3′, 5′-CGGGCAGCCTGTGAGACCTTTGA-3′, mu-phox-p47: 5′-GAGGCGGAGGATCCGG-3′, 5′-TCTTCAACAGCAGCGTACGC-3′, 5′-CAACTACGCAGGTGAACCGTATGTAACCATCA-3′, mu-TNF-α: 5′- TTCTATGGCCCAGACCCTCA-3′, 5′-TTGCTACGACGTGGGCTACA-3′, 5′- CTCAGATCATCTTCTCAAAATTCGAGTGACAAGC-3′, mu-IL-6 : 5 ′-TGTTCTCTGGGAAATCGTGGA-3 ′, 5 ′-AAGTGCATCATCGTTGTTCATACA-3′, 5′-ATGAGAAAAGAGTTGTGCAATGGCAATTCTG-3′, mu-IL-10 : 5′-CCAGAGCCACATGCTCCTAGA-3 ′, 5 ′-TGGTCCTTTGTTTGAAAGAAAGTCT-3′, 5′- TGCGGACTGCCTTCAGCCAGG-3′, mu-IL-18: 5 ′-GACTCTTGCGTCAACTTCAAGGA-3 ′, 5 ′-TTGTCTGATTCCAGGTCTCCATT-3′, 5′-TGATGTTTATTGACAACACGCTTTACTTTATACCTGAAGA-3′, mu-IL-23-p19: 5 ′-AGCGGGACATATGAATCTACTAAGAGA-3 ′, 5 ′-GTCCTAGTAGGGAGGTGTGAAGTTG-3′, 5′- CCAGTTCTGCTTGCAAAGGATCCGC-3′ mu-IL-23-p40: 5 ′-GACCATCACTGTCAAAGAGTTTCTAGAT-3 ′, 5 ′-AGGAAAGTCTTGTTTTTGAAATTTTTTAA-3′, 5′-CCACTCACATCTGCTGCTCCACAAGAAG-3′, mu-Mcp-1: 5′-CTTCTGGGCCTGCTGTTCA-3′, 5′-CCAGCCTACTCATTGGGATCA-3′, 5′-CTCAGCCAGATGCAGTTAACGCCCC-3′, mu-Mip-1α: 5′-AGCTGACACCCCGACTGC-3′, 5′-GTCAACGATGAATTGGCGTG-3′, 5′-TGCTGCTTCTCCTACAGCCGGAAGAT-3′, mu-18S rRNA: 5′-CCTGCCCTTTGTACACACCG-3′, 5′-CGATCCGAGGGCCTCAC-3′, 5′-CCGTCGCTACTACCGATTGGATGGTTT-3′. Unless otherwise indicated, qPCR data are expressed as mean ± SD of five independent experiments.

Immunoblot analysis

Protein extracts from RAW-21/37 cells were prepared using the radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris [tris(hydroxymethyl)aminomethane] HCl, pH 8.0;) supplemented with 1 μg/ml aprotinin and 1 μg/ml leupeptin (all from Sigma). 10 – 40 μg of total protein were run on 8 –15% SDS-polyacrylamide gels and immunoblotting was performed exactly as described (Theurl et al., 2006) using a mouse anti-human TfR1 antibody (1:1000; Zymed), a rabbit anti-human ferritin antibody (1:800; Dako), a rabbit anti-human ferroportin 1 antibody (1:400; Eurogentec), a rabbit anti-Slc11a1/Nramp1 antibody (1:1000; (Atkinson et al., 1999)), a rabbit anti-rat Hmox1 (1:2000; Stressgen) or a rabbit anti-Actin antibody (1:1000; Sigma), the latter used as a loading control.

Quantification of iron uptake and release by macrophages

For macrophage iron uptake and release studies, RAW-21/37 cells were infected with S. Typhimurium as detailed above. Following three washing cycles with serum-free, HEPES-buffered DMEM (2 mM L-glutamine, 50 μg/ml ampicillin, 25 mM HEPES, pH 7.4), cells were then incubated therein. For determination of the uptake of non-transferrin-bound iron (NTBI), 5 μM ⁵⁹Fe-citrate (DuPont New England Nuclear) were used, whereas for the measurement of transferrin-bound iron (TBI) acquisition, ⁵⁹Fe-labeled transferrin was added at a concentration of 12.5 μg/ml exactly as described (Ludwiczek et al., 2003). For iron release experiments, cells were first incubated with 5 μM ⁵⁹Fe-citrate for 4 hours to allow for iron loading and then washed four times and repleted with 2 ml of serum-free, HEPES-buffered DMEM. After an additional incubation for 2 hours, cellular iron release was measured by means of a γ-counter as described (Nairz et al., 2008). In parallel to each iron release study, a trypan-blue exclusion assay was performed to ensure that neither treatment interfered with the integrity of the RAW-21/37 cell surface membrane.

Determination of iron acquisition by S. Typhimurium

RAW-21/37 cells were seeded in complete DMEM and infected with S. Typhimurium strain C5RP4 at an MOI of 10 as detailed above. Thereafter, cells were washed three times and repleted with serum-free, HEPES-buffered DMEM. After the addition of either 5 μM ⁵⁹Fe-citrate or 12.5 μg/ml of ⁵⁹Fe-labeled transferrin, macrophages were incubated for additional 4 hours. Thereafter, intracellular bacilli were harvested according to a modified protocol as described (Olakanmi et al., 2002). Briefly, RAW-21/37 cells were lysed in 0.1% SDS in the presence of 10,000 U/ml DNase-I (Invitrogen) and EDTA-free protease inhibitor (Roche). Cell lysates were withdrawn into microfuge tubes and incubated at room temperature for 10 min, and aliquots of 50 μl volume were transferred into γ-counter tubes for assessment of total iron in the macrophage lysate. The released bacilli were centrifuged at 10,000 g for 10 min at room temperature. The supernatant was removed, and the bacterial pellet was washed three times with 0.01% SDS in DMEM containing 1 mg/ml proteinase K (Roche). The bacilli were finally resuspended in 500 μl of 0.01% SDS in DMEM containing 1 mg/ml proteinase K. Fifty μl of the bacterial suspension were plated in serial dilutions onto agar plates to quantify released bacteria while the remaining volume was filtered through centrifugal filter devices with a PDVF membrane of 0.22 μm pore size (Millipore). The filters containing the trapped bacilli were washed five times and placed into a γ-counter tube. The amount of Salmonella-associated ⁵⁹Fe was assessed using a γ-counter. No association of ⁵⁹Fe to S. Typhimurium, heat-inactivated at 70°C for 20 min, could be detected. The incorporation of ⁵⁹Fe into ferritin multimers was investigated exactly as described REF.

Detection of cytokines, ferritin content, nitrite and total oxidative capacity

Concentrations of TNF, IL-1β, IL-6 and IL-10 in cell culture supernatants were determined by specific ELISA kits (BD Pharmingen). Ferritin content in cytoplasmatic extracts was measured by a specific ELISA kit (Kamiya Biomedical Company) and normalized for total cytoplasmatic protein content as determined by the BCA method (Pierce). Determination of nitrite, an oxidation product of NO, was carried out to monitor the production of RNS using the Griess reagent (Merck) as described (Fritsche et al., 2001). The formation of ROS was studied by measuring total oxidative capacity (TOC) in culture supernatants with the TOC Kit (Tatzber KG), which is based on the peroxidase-mediated oxidation of tetramethylbenzidine (Weiss et al., 2003).

Electrophoretic mobility shift assays (EMSA)

Extraction of nuclear proteins was carried out as described (Schreiber et al., 1989). For the preparation of radiolabeled probes representing standard consensus sequences of various transcription factors known to mediate the response of murine macrophages to LPS/IFN-γ stimulation, the oligonucleotides used were as follows: Interferon regulatory factor 1 (IRF-1) sense 5′-GGAAGCGAAAATGAAATTG-3 ′, IRF-1 antisense 5′-TGAGTCAATTTCATTTTCG-3′; NF-κB sense 5′-AGCTTCAGAGGGGACTTTCCGAGAGG-3′, NF-κB antisense 5′-TCGACCTCTCGGAAAGTCCCCTCTGA-3 ′ ; STAT-1 sense 5 ′-CATGTTATGCATATTCCTGTAAGTG-3 ′, STAT-1 antisense 5 ′-CGTGCACTTACAGGAATATGCATA-3 ′ ; STAT-3 sense 5′-GATCCTTCTGGGAATTCCTA-3′, STAT-3 antisense 5′-GATCTAGGAATTCCCAGAAG-3′. For the preparation of double-stranded probes, oligomers were annealed, and overhanging ends were filled with [α-32P]dCTP (Amersham) and the three other non-radiolabeled dNTPs (Pharmacia) using Klenow enzyme (Amersham) as described (Fritsche et al., 2003). For the preparation of unlabeled competitors, dCTP was used. A total of 10 μg of nuclear extracts were preincubated with 2 μg of double-stranded poly(dI-dC) · poly(dIdC) (Pharmacia) on ice for 10 min before addition of 2 ng of the radiolabeled oligonucleotide probe (50,000 cpm/ng). For competition studies, a 30-fold excess of unlabeled oligonucleotide probe was added to the nuclear extracts derived from RAW-21 cells (stimulated with S. Typhimurium and IFN-γ for 24 hours) 10 min before addition of the radioactive probe. The DNA binding reactions were performed in the presence of 200 mM HEPES (pH 7.8), 10 mM EDTA, and 10 mM DDT for 20 min on ice. Samples were separated on 6% nondenaturing polyacrylamide gels, which were subsequently exposed by means of autoradiography.

Statistical analysis

Statistical analysis was carried out using a SPSS statistical package. Calculations for statistical differences were carried out by parametric ANOVA test using Bonferroni correction for multiple tests or, where appropriate, by nonparametric Kruskal-Wallis test. Unless otherwise indicated, statistically significant differences between means are indicated as follows: * P < 0.05 for infected/stimulated RAW-21 macrophages in comparison to RAW-21 controls; ° P < 0.05 for infected/stimulated RAW-37 in comparison to untreated RAW-37 cells; # P < 0.05 for the comparison of RAW-21 and RAW-37 cells subjected to the same treatment.

Supplementary Material

Supp Data

Supplementary figure S1. Effects of Slc11a1 on non-transferrin-bound iron acquisition by Salmonella-infected macrophages Slc11a1-functional RAW-37 cells (dark grey bars) and Slc11a1-non-functional RAW-21 cells (light grey bars) were infected with S. Typhimurium (S. tm.) at a MOI of 10 and stimulated with 50 U/ml IFN-γ for 24 hours. Dmt1 mRNA levels (Supplementary figures 1A) were determined by qPCR. Values were corrected for the amount of 18S ribosomal RNA, which was determined in parallel. Results are shown as relative differences of this ratio in comparison to unstimulated Slc11a1-expressing RAW-37 control macrophages (=1.0). Data are expressed as mean ± SD of five independent experiments. The uptake of NTBI (depicted in supplementary figure 1B) was determined as described in ‘Experimental procedures’. Data are shown as mean ± SD of five independent experiments performed in duplicates and are expressed as fold change in the relative iron uptake in comparison to RAW-37 controls (=1.0). * P < 0.05 for infected/stimulated RAW-21 macrophages in comparison to RAW-21 controls; ° P < 0.05 for infected/stimulated RAW-37 in comparison to untreated RAW-37 cells; # P < 0.05 for the comparison of RAW-37 and RAW-21 cells subjected to the same treatment.

Supplementary figure S2. Impact of Slc11a1 on macrophage immune effector functions in the presence of IFN-γ stimulation RAW-37 and RAW-21 cells were infected with Salmonella for 24 hours and activated with IFN-γ. The mRNA levels of iNOS were determined by qPCR (Supplementary figure 2A). Values were corrected for the amount of 18S ribosomal RNA, determined in parallel. Results are shown as relative differences of this ratio in comparison to the unstimulated control macrophages (=1.0). Cell culture supernatants were analyzed for the formation of ROS (as determined by measuring the total oxidative capacity (TOC), supplementary figure 2B) and of IL-1β (Supplementary figure 1C) as described in ‘Experimental procedures’. Statistically significant differences are indicated as described in the legend to figure 1.

Supplementary figure S3. Binding affinities of key macrophage transcription factors in Salmonella-infected RAW-21 and RAW-37 cells RAW-37 and RAW-21 cells were infected with S. Typhimurium and treated with solvent or stimulated with IFN-γ for 3 or 24 hours. Nuclear proteins were prepared and analyzed for the binding activities of NF-κB (upper panel), STAT-1 (middle panel) and STAT-3 (lower panel) in EMSAs. The specificity of binding was confirmed by cold competition using extracts of RAW-21 cells infected and stimulated for 24 hours and a 30-fold excess of the same unlabeled oligonucleotide (cold probe, far right lane). One of four representative experiments is shown.

Supplementary figure S4. IL-10 production in RAW-37 and RAW-21 cells infected with iron-uptake mutant Salmonella RAW-37 and RAW-21 phagocytes were infected with S. Typhimurium wild-type strain as well as with isogenic entC and entC sit feo mutant derivatives and treated with PBS or 12.5 μg/ml holo-transferrin. IL-10 concentrations in culture supernatants were determined by a specific ELISA.

Supplementary figure S5. Stable expression of Slc11a1 in solvent-treated and Salmonella-infected RAW-37 Slc11a1-expressing RAW-37 cells and RAW-21 cells were infected with Salmonella and stimulated with IFN-γ exactly as described in the legend to Figure 1. Whole cell lysates were analysed by immunoblotting using specific antibodies to Slc11a1 (upper panel) and to actin (lower panel). One of four representative immunoblot experiments is shown.