Siderocalin is a secreted protein that binds to siderophores to prevent bacterial iron acquisition. While it has been shown to inhibit the growth of Mycobacterium tuberculosis (M.tb) in extracellular cultures, its effect on this pathogen within macrophages is not clear. Here, we show that siderocalin expression is upregulated following M.tb infection of mouse macrophage cell lines and primary murine alveolar macrophages. Furthermore, siderocalin added exogenously as a recombinant protein or overexpressed in the RAW264.7 macrophage cell line inhibited the intracellular growth of the pathogen. A variant form of siderocalin, which is expressed only in the macrophage cytosol, inhibited intracellular M.tb growth as effectively as the normal, secreted form, an observation that provides mechanistic insight into how siderocalin might influence iron acquisition by the bacteria in the phagosome. Our findings are consistent with an important role for siderocalin in protection against M.tb infection and suggest that exogenously administered siderocalin may have therapeutic applications in tuberculosis.
Mycobacterium tuberculosis (M.tb) is a bacterial pathogen that leads to the death of approximately 2 million people every year (Kaufmann & McMichael, 2005). Following inhalation, M.tb is taken up by macrophages and other cells in the lung, and can survive intracellularly for extended periods (Kaufmann, 2005). Iron is essential for M.tb to survive and grow intracellularly and to evade host immune defenses (Raghu et al., 1993; Kelley & Schorey, 2003; Serafin-Lopez et al., 2004). To ensure that its demand for iron is met, M.tb synthesizes and secretes the salicylate-type siderophore carboxymycobactin that tightly binds iron and delivers it to the bacteria via specific surface receptors and transport proteins (Ratledge & Dover, 2000). Mycobacterium tuberculosis also expresses a more hydrophobic form of carboxymycobactin known as mycobactin that remains associated with the bacterial cell wall rather than being secreted (Rodriguez, 2006). Its role in iron acquisition is not clear, but it may facilitate the action of carboxymycobactin. Strains of M.tb that are unable to produce or utilize carboxymycobactin grow poorly under iron-restricted conditions, and exhibit attenuated growth inside macrophage cell lines and in vivo in mouse models of tuberculosis, confirming a role for siderophore-mediated iron acquisition in M.tb growth and virulence (De Voss et al., 2000; Rodriguez, 2006).
The host immune system has developed multiple strategies to limit the availability of iron to microbial pathogens (Doherty, 2007). One of these strategies involves the protein siderocalin [also known as lipocalin 2, 24p3 and neutrophil gelatinase-associated lipocalin (NGAL)] (Goetz et al., 2002; Flo et al., 2004; Berger et al., 2006), a 21-kDa member of the lipocalin superfamily that was originally identified as a constituent of specific granules in human neutrophils (Kjeldsen et al., 1993), and later shown to be an acute-phase protein secreted by the liver in mice (Liu & Nilsen-Hamilton, 1995). Siderocalin does not bind iron directly (Yang et al., 2002), but rather, it binds specifically to iron-laden bacterial siderophores, including enterobactin secreted by Escherichia coli (Goetz et al., 2002) and carboxymycobactin produced by mycobacteria (Holmes et al., 2005). The addition of recombinant siderocalin to cultures of E. coli or M.tb in vitro has a bacteriostatic effect that can be reversed with iron supplementation (Goetz et al., 2002; Flo et al., 2004; Martineau et al., 2007). Furthermore, mice with a targeted disruption of the siderocalin gene exhibited a higher mortality rate than wild-type animals when challenged with intraperitoneal injections of pathogenic E. coli (Flo et al., 2004; Berger et al., 2006). These observations point to an important role for siderocalin in antibacterial defense.
Although siderocalin has been shown to inhibit the growth of M.tb in extracellular culture, its effect on the growth of the pathogen in macrophages remains unclear. To clarify this issue, we have carried out experiments to examine the effect of siderocalin on the intramacrophage multiplication of M.tb.
Materials and methods
Mycobacterium tuberculosis infection of mice
Female C57BL/6J mice (Jackson Labs) were housed under specific pathogen-free conditions at the Harvard School of Public Health and were provided chow and water ad libitum. Following 3 weeks of acclimation in the BSL3 facility, 7–9-week-old mice were infected with virulent M.tb (Erdmann strain; Trudeau Institute) via aerosol. Noninfected mice housed under the same conditions served as controls. Alveolar macrophages were prepared following a standard protocol (Zhang et al., 2008). Lungs were harvested from the mice at the time of sacrifice.
RNA isolation, cDNA synthesis, and quantitative reverse transcriptase (qRT)-PCR
Total RNA was prepared from whole lung tissue, alveolar macrophages, and RAW264.7 cells using TRIZOL (Invitrogen) according to the manufacturer's specifications. Gene expression was analyzed by qRT-PCR using the SYBR GreenER qPCR SuperMix for ABI PRISM (Invitrogen). Acidic ribosomal phosphoprotein P0 (Arbp, also known as 36B4) or GAPDH served as reference genes. The primer sequences for siderocalin were as follows: fwd, 5′-CCCCTGAACTGAAGGAACGTT-3′, rev, 5′-CGTCCTTGAGGCCCAGAGA-3′. Calculations for relative quantification were performed using the comparative CT method ().
Generation of a cytosolic mutant of human siderocalin
The primers, ngalF, 5′-GGATCCGCCACCATGCAGGACTCCACCTCAGACCTG-3′, and ngalR, 5′-GAATTCTCAGCCGTCGATACACTG-3′, were used to PCR amplify a 557-bp fragment using the plasmid pDNA3.1NGAL containing the full-length human siderocalin cDNA (Bundgaard et al., 1994) as the template. The amplification product, which retained the methionine start codon but lacked the secretion signal sequence, was cloned into the pCDNA3.1 vector to yield the construct pCDNA3.1cytoNGAL. Expression analysis was initially performed following transient transfection into HeLa cells using Lipofectamine 2000 (Invitrogen). The cells were processed for Western blotting or immunofluorescence with an anti-human siderocalin monoclonal antibody (mAb) (Kjeldsen et al., 1996) 48 h after transfection.
Construction and purification of recombinant glutathione-S-transferase (GST)–siderocalin
The insert from pCDNA3.1cytoNGAL was cloned into the vector pGEX2TK to generate the plasmid pGEXcytoNGAL, encoding an inframe fusion with GST. pGEXcytoNGAL was transformed into E. coli strain BL21 to avoid binding of endogenous bacterial siderophores (Goetz et al., 2002). The GST–siderocalin fusion protein and GST were affinity purified from the relevant transformants using glutathione-agarose as described previously (Chaudhuri et al., 1997). The purity and identity of the fusion protein were confirmed by Western blotting with an anti-human siderocalin mAb (Kjeldsen et al., 1996).
Escherichia coli and M.tb growth assays
The ability of GST–siderocalin to inhibit bacterial growth was confirmed using the siderocalin-susceptible E. coli strain H9049 (kindly provided by Dr Michael Fischbach, Harvard Medical School) and M.tb in previously described in vitro growth inhibition assays (Fischbach et al., 2006; Martineau et al., 2007).
Mycobacterial infection of murine macrophages
Mycobacterium tuberculosis H37Rv (Wolf et al., 2007) and Mycobacterium bovis-bacillus Calmette-Guérin (BCG) (Teitelbaum et al., 1999) expressing green fluorescent protein (GFP) were cultured in 7H9 medium containing 50 µg mL−1 kanamycin (Sigma). Mycobacterium tuberculosis and BCG in midgrowth phase were pelleted, resuspended in complete Dulbecco's modified Eagle's medium (DMEM), sonicated, and passed through a 5-µM filter. For qRT-PCR analysis, RAW264.7 macrophages were infected at a multiplicity of infection (MOI) of 5 : 1 for the indicated time points and harvested in TRIZOL. For intracellular CFU, RAW264.7 macrophages were infected at an MOI of 1 : 1 to avoid artifacts associated with bacterial clumping. To analyze the effect of recombinant siderocalin on intracellular M.tb and BCG growth, GST or GST–siderocalin was added to complete DMEM at a concentration of 10 µg mL−1. At the indicated time points, infected cells were washed with sterile phosphate-buffered saline (PBS), lysed using 1% Trition X-100 in PBS, serially diluted and plated on 7H11 agar. Infected cells incubated with GST or GST–siderocalin grown on glass coverslips were fixed with 3.7% paraformaldehyde and mounted with Dako fluorescent mounting medium for examination by fluorescence microscopy.
Generation of RAW264.7 cells stably expressing wild-type or cytosolic siderocalin
RAW264.7 cells were transfected with pCDNA3.1NGAL or pCDNA3.1cytoNGAL using Lipofectamine 2000. Stable transfectants were selected and grown in 400 µg mL−1 of G418.
Cells lysates were subjected to electrophoresis and Western blotting with the antisiderocalin mAb specific for either the mouse (R&D Systems) or human (Kjeldsen et al., 1996) protein and developed with a horseradish peroxidase-conjugated anti-mouse IgG antibody and enhanced chemiluminescence (Pierce).
Cells plated on coverslips were fixed, permeabilized, and stained with antisiderocalin, followed by fluorochrome-conjugated secondary antibody. The cells were mounted in Vectashield before examination by fluorescence microscopy.
For statistical analysis, the means±SEM are displayed. For intracellular M.tb growth, statistical analysis was performed using two-way anova. The Student t-test was utilized for all other statistical comparisons. All statistical functions were executed using graphpad prism version 4.0 for Windows (GraphPad Software).
Mycobacterium tuberculosis infection induces a significant increase in siderocalin expression in murine macrophages and lung tissue
Because it is well known that M.tb targets host macrophages in order to establish infection (Kaufmann, 2004), we investigated the expression of siderocalin in RAW264.7 murine macrophages. Siderocalin was significantly upregulated at 24 h postinfection, exhibiting an approximately 300-fold induction when compared with noninfected controls (Fig. 1a). This increase in siderocalin mRNA was accompanied by a corresponding increase in siderocalin protein expression and secretion (Fig. 1b). Similar results were obtained with the J774 murine macrophage cell line (data not shown).
To examine expression of siderocalin in vivo, the lungs of mice infected with virulent M.tb were isolated and subjected to qRT-PCR analysis. Although the level of siderocalin expression in total lung tissue was much less than in the macrophage cell lines, there was a reproducible and statistically significant increase at 12 h postinfection, which returned to the baseline by 24 h (Fig. 2a). We also examined siderocalin mRNA levels in alveolar macrophages from control and infected mice and found a more robust increase than in whole lung, although the time course of expression was similar (Fig. 2b).
Recombinant siderocalin significantly inhibits M.tb growth within murine macrophages
To address the impact of siderocalin on the growth of M.tb within macrophages, we prepared a recombinant GST–siderocalin fusion protein (Fig. 3a) and confirmed its ability to inhibit the growth of a siderocalin-susceptible strain of E. coli (Fig. 3b). When added directly to the bacterial culture, recombinant siderocalin also significantly inhibited growth of virulent H37Rv M.tb in a dose-dependent manner (Fig. 4a) and at levels consistent with previous studies (Martineau et al., 2007). Because 10 µg mL−1 of siderocalin had the highest effect in these experiments, this concentration of protein was used to treat RAW264.7 cells infected with M.tb. Figure 4b demonstrates that addition of recombinant siderocalin into the macrophage culture medium resulted in a significant reduction of intracellular M.tb CFU over a 5-day time course, but had no effect on the intracellular growth of BCG (Fig. 4c).
Overexpression of wild-type and a cytosolic form of siderocalin significantly reduces intracellular M.tb proliferation
Although both siderocalin and M.tb carboxymycobactin are secreted (Liu & Nilsen-Hamilton, 1995; Holmes et al., 2005), the location in which these molecules interact in the context of intramacrophage infection is not well understood. In order to investigate this issue, we constructed a variant of siderocalin that lacked the signal sequence required for secretion. When transiently overexpressed in HeLa cells, this variant was present in whole cell lysates, but unlike wild-type siderocalin, it could not be detected in the cell culture supernatant (Fig. 5a). Immunofluorescence analysis of transiently transfected HeLa cells and stably transfected RAW264.7 macrophages demonstrated a reticular and punctate distribution of the wild-type siderocalin, consistent with localization to the secretory pathway. On the other hand, the cytosolic variant demonstrated a diffuse staining pattern characteristic of soluble cytoplasmic proteins (Fig. 5b). Infection of RAW264.7 cells stably expressing wild-type or cytoplasmic siderocalin demonstrated that both proteins had similar bacteriostatic effects on the intracellular growth of M.tb (Fig. 6).
Intracellular M.tb undergoes growth arrest following addition of recombinant siderocalin
In order to determine whether the addition of siderocalin affected intracellular growth or cell-to-cell spread of M.tb, RAW264.7 cells infected with GFP-M.tb and incubated with recombinant siderocalin were subjected to immunofluorescence analysis. Although no difference could be detected in the number of infected cells, there were distinct morphological changes observed in the bacteria present in the cells treated with recombinant siderocalin. Three days postinfection, M.tb from control cells increased from one to numerous small bacilli (Fig. 7, upper panels). However, M.tb from siderocalin-treated cells did not exhibit an obvious increase in the number of bacilli. Instead, only one or two elongated, filamentous bacilli were detected (Fig. 7, lower panels), a morphologic appearance consistent with bacteria that are growth arrested (Liu & Nilsen-Hamilton, 1995; Hett & Rubin, 2008; Hett et al., 2008).
The observations presented here indicate that siderocalin expression is upregulated by M.tb in mouse macrophage cell lines after infection in vitro and in alveolar macrophages following aerosol infection of mice in vivo. Moreover, our results demonstrate clearly that siderocalin inhibits M.tb multiplication both extracellularly, as has been shown previously (Martineau et al., 2007), and also when the bacteria are growing in macrophages. Similar findings, based on both in vitro and in vivo experiments, have been reported recently by others (Saiga et al., 2008). However, in that study, the siderocalin-dependent inhibition of intracellular growth of M.tb and BCG was apparent only in alveolar epithelial cells and not alveolar macrophages. The reason for the difference between our findings and those of Saiga (2008) with respect to the effects of siderocalin on growth of M.tb in macrophages is not clear. It could relate to the fact that we used a macrophage cell line while they used primary alveolar macrophages. The difference would then suggest that the effects of siderocalin on intracellular bacterial growth are influenced by the characteristics of specific host cell types.
Our data indicate that siderocalin does not simply act on extracellular M.tb during the course of cell-to-cell spread, but actually limits the replication of bacteria inside the macrophages. This result implies that siderocalin must somehow influence iron acquisition by M.tb residing within the macrophage. One possibility is that siderocalin is internalized by receptor-mediated endocytosis into the intracellular compartment in which the bacteria reside. We have looked for evidence in support of this idea, but unlike results obtained by others with alveolar epithelial cells (Saiga et al., 2008), we have not been successful in colocalizing overexpressed or exogenously administered siderocalin with M.tb residing within RAW264.7 macrophages (C.V. Srikanth & E.E. Johnson, unpublished data). Another potential explanation is that siderocalin may reduce the concentration of intracellular free iron via interactions with its putative receptor, as has been suggested previously (Devireddy et al., 2005). The mechanism of this effect is not well understood, but it is believed to involve a hypothetical mammalian siderophore that shuttles iron into and out of the cell. Because intracellular iron is utilized by M.tb (Olakanmi et al., 2002, 2006), siderocalin-mediated depletion of this nutrient source offers a plausible explanation for the ability of siderocalin to inhibit the intramacrophage growth of the pathogen. However, we have not observed any consistent effects of exogenously added siderocalin on intracellular free iron concentrations in RAW264.7 cells based on a calcein fluorescence assay (Wang et al., 2008; L. Wang, unpublished data). A third possibility is that siderophores produced by intracellular M.tb may traverse the phagosomal or plasma membranes to scavenge iron from both cytoplasmic and extracellular sites, and that extracellular siderocalin may interfere with this process by sequestering the siderophores without necessarily accessing the compartment containing the bacteria or altering intracellular iron stores. This idea is consistent with the relatively hydrophobic nature of carboxymycobactin (Holmes et al., 2005; Luo et al., 2005), and is also supported by our finding that a cytosolic form of siderocalin is as effective at inhibiting the intracellular growth of M.tb as the normal, secreted form (Fig. 6).
What is the reason for the differential susceptibility of M.tb and BCG to siderocalin observed in our experiments (Fig. 4b and c)? An analysis of siderophores secreted by various mycobacterial species grown under iron-limiting conditions may shed some light on this question. This study demonstrated that BCG produced significantly lower amounts of carboxymycobactin than M.tb and also expressed certain carboxymycobactins with side chain structures not found in M.tb (Gobin et al., 1999). These differences raise the possibility that BCG carboxymycobactin may be relatively resistant to inhibition by siderocalin, particularly because the structure of the carboxymycobactin side chain is known to influence binding to siderocalin (Holmes et al., 2005). It is also possible that M.tb and BCG may reside in intracellular locations that are differentially accessible to siderocalin. It was recently reported that M.tb and Mycobacterium leprae are able to escape endomembranous compartments and enter the cytosol, while BCG remains restricted to the phagosome (van der Wel et al., 2007). Differences in subcellular localization between M.tb and BCG may contribute to the differences in their susceptibility to siderocalin. Cytosolic residence of M.tb may also account for our observation that a cytosol-restricted variant of siderocalin was able to inhibit growth of this pathogen (Fig. 6).
Competition for iron between host and pathogen is crucial to the outcome of infection (Schaible & Kaufmann, 2004; Doherty, 2007). One of the mechanisms used by the host to deprive pathogens of iron is production of hepcidin, an acute-phase response protein that decreases expression of the macrophage and enterocyte iron exporter ferroportin, thereby blocking the release of iron into the circulation (Nicolas et al., 2002; Nemeth et al., 2003, 2004). While this mechanism may be effective against extracellular microorganisms, it may actually favor the growth of pathogens such as M.tb that reside within macrophages, because hepcidin-mediated downregulation of ferroportin would increase intracellular iron levels in these cells. An increase in siderocalin expression, which occurs locally and systemically during the response to infection (Liu & Nilsen-Hamilton, 1995; Cowland et al., 2003; Saiga et al., 2008), may help to counteract the effects of hepcidin by disrupting iron acquisition by intracellular pathogens.
Siderocalin is emerging as a significant factor in implementing the strategy of iron deprivation that is an important component of antimicrobial defense. Our findings, together with observations reported recently by others (Martineau et al., 2007; Saiga et al., 2008), highlight the role of this molecule in protection against M.tb, and raise the possibility that exogenously administered siderocalin may find use as an adjunct to standard chemotherapy in the treatment of tuberculosis.
The authors are grateful to Dr Michael Fischbach for providing the H9049 bacterial strain, and to Dr Marianne Wessling-Resnick for encouragement and advice. This work was supported by grants to B.J.C. from the NIH (R21 AI065619) and the Milton Fund of Harvard University. E.E.J. was supported by an NIH Roadmap Grant to Dr Marianne Wessling-Resnick and L.W. by an unrestricted grant from Wyeth Nutrition.