Abstract
Natural-resistance associated macrophage protein 1 (Nramp1) encodes a transmembrane phagolysosomal protein exerting resistance toward infections with intracellular pathogens by a mechanism not fully elucidated so far. We used the murine macrophage cell line RAW264.7, stably transfected with functional (RAW-37) or nonfunctional (RAW-21) Nramp1, to study for differences in the expression of NO, a central antimicrobial effector molecule of macrophages. Following stimulation with IFN-γ and LPS, Nramp1-expressing cells exhibit higher enzymatic activity of inducible NO synthase (iNOS) and increased cytoplasmic iNOS mRNA levels than RAW-21 cells. Time-course experiments showed that iNOS-mRNA levels remain increased in RAW-37 cells after prolonged cytokine stimulation while they decrease in RAW-21 cells. Reporter gene assays with iNOS-promoter luciferase constructs demonstrated an increased and prolonged promoter activity in Nramp1-resistant vs susceptible cells. This was paralleled by increased IFN regulatory factor 1 (IRF-1) expression and binding affinity to the iNOS promoter in RAW-37 cells, which may be related to enhanced STAT-1 binding affinity in these cells. A point mutation within the IRF-1 binding site of the iNOS promoter abolished the differences in iNOS transcription between RAW-21 and RAW-37 cells. Cells carrying functional Nramp1 express increased amounts of NO, which may be related to STAT-1-mediated stimulation of IRF-1 expression with subsequent prolonged activation of iNOS transcription. Enhanced NO expression may partly underlie the protection against infection with intracellular pathogens by Nramp1 functionality.
In mice, resistance to infection with various intramacrophage pathogens such as Salmonella typhimurium, Leishmania donovani, Mycobacterium bovis, and Mycobacterium lepraemurium is controlled by a single dominant gene originally designated Bcg/Ity/Lsh (Refs. 1, 2, 3 , respectively) located on mouse chromosome 1 (reviewed in Refs. 4, 5, 6). A candidate gene for Bcg was identified by positional cloning and termed natural resistance associated macrophage protein 1 (Nramp1)2 (7). Meanwhile, it was also called solute carrier family 11 member 1. Murine Nramp1 is expressed exclusively in phagocytic cells and encodes a highly glycosylated 548 aa integral membrane protein with 10–12 putative transmembrane-spanning domains (7, 8). It is recruited to the membrane of late phagosomes (9) where it is supposed to function as a transporter of divalent cations such as Fe2+, Mn2+, and Zn2+ (10). Interestingly, resistance to infection with intracellular parasites is abrogated by a single glycine-to-aspartic amino acid substitution at position 169 (G169D) within putative transmembrane domain 4 (11).
However, the exact biochemical function of Nramp1 is still far from being clear. Some reports consider Nramp1 to transport iron into phagolysosomes (12, 13, 14) where it could induce enhanced formation of reactive oxygen intermediates via the Fenton and Haber-Weiss reactions (15), which may be further enhanced by the putative ability of Nramp1 to accumulate protons within the late phagolysosome. Other data support a role for Nramp1 in iron efflux out of phagolysosomes (16, 17) and finally even out of the macrophages (18, 19), thus depriving microorganisms of the essential growth factor iron.
Intracellular iron availability is well known to influence the expression of inducible NO synthase (iNOS). Moreover, because Nramp1 may transport iron, it may thereby modulate iNOS expression by a transcriptional mechanism (20) involving iron-mediated deactivation of critical transcription factors such as NF-IL6 (C/EBP-β) (21). iNOS catalyzes the formation of NO, which is a central effector molecule in antimicrobial activity exerted by macrophages stimulated with various pro-inflammatory cytokines such as IFN-γ, LPS, TNF-α, and IL-1β (22, 23). Cells carrying functional Nramp1 exhibit an increased production of NO (24, 25, 26). The importance of NO generation and its modulation by Nramp1 for host defense against intracellular pathogens has been demonstrated by the observation that growth inhibition of mycobacteria by Nramp-resistant cells can be abrogated by the addition of iNOS inhibitors (27). Nramp1 was supposed to stabilize the mRNA of several IFN-γ-inducible genes, such as TNF-α (28), and TNF-α seems to be an important costimulus for NO production, as TNF-α knockout macrophages infected with S. typhimurium show no iNOS expression (29). Thus, the aim of our study was to examine the effect of Nramp1-functionality on iNOS expression and to elucidate the underlying molecular mechanism.
Materials and Methods
Cell culture techniques
Murine RAW264.7 macrophages, which are derived from BALB/c mice and thus Nramp-susceptible, were stably transfected with the Nramp1 gene in sense (RAW-37) or antisense (RAW-21) orientation as described previously (30). Nramp1 polypeptide expression in RAW-37 cells was confirmed by Western blotting and found to be absent in RAW-21 cells. Cells were grown in DMEM (Biochrom, Berlin, Germany) supplemented with 10% FBS, 2 mM l-glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin and periodically selected for plasmid retention by growth in DMEM containing 1 mg/ml G418 (Invitrogen, Lofer, Austria). Cells were stimulated with recombinant murine IFN-γ (Coachrom, Vienna, Austria) and LPS (Sigma-Aldrich, Munich, Germany).
For coinfection experiments, the S. typhimurium strain C5RP4 (kindly provided by P. Mastroeni, Cambridge University, Cambridge, U.K.) was used. Before infection of RAW macrophages, the bacteria were grown overnight in Luria Bertoni medium supplemented with 50 μg/ml ampicillin (Sigma-Aldrich). A total of 2 × 106 RAW-21/37 cells were seeded into 6-well plates in 2 ml of DMEM supplemented with 10% FBS, 2 mM l-glutamine, and 50 μg/ml ampicillin. After 12 h, cells were infected with a 10-fold amount of Salmonella (preincubated in DMEM for 30 minutes) for 50 min at 37°C. Then, cells were washed three times with PBS and incubated in 2 ml of DMEM supplemented with 20 μg/ml gentamycin (Sigma-Aldrich) before stimulation with murine IFN-γ. After 24 h, the concentration of nitrite in tissue culture supernatants was determined by the Griess reaction (31) using Griess-Ilosvays reagent from Merck (Darmstadt, Germany). Briefly, 250 μl of sample or sodium nitrite standard were mixed with 500 μl of Griess reagent and incubated at room temperature for 15 min before OD was measured photometrically at 546 nm.
Northern blot analysis
Cells were harvested by scraping and were washed twice with PBS before total RNA was isolated using RNA-Clean (AGS, Heidelberg, Germany). Ten micrograms of total RNA were separated on 1% agarose/2.2 M formaldehyde gels and then transferred to Duralon-UV membranes (Stratagene, La Jolla, CA) by capillary blotting. After UV cross-linking and prehybridization for 6–8 h at 65°C, blots were hybridized overnight with [α-32P]d cytidine 5′-triphosphate (dCTP)-radiolabeled cDNA probes at 65°C. The hybridization solution contained 3× SSC, 0.1% SDS, 0.1% sodium pyrophosphate, 10% dextran sulfate, 10× Denhardt’s solution (0.2% Ficoll 400, 0.2% polyvinylpyrrolidone, and 0.2% BSA), and 1 mg/ml denatured salmon sperm DNA. After hybridization the filters were washed and then exposed for autoradiography.
iNOS enzyme activity assay
Measurement of iNOS enzyme activity was performed in 96-well microtiter plates as follows (32): 30 μg of macrophage lysate were incubated in 20 mM Tris-HCl (pH 7.9) in the presence of 3 mM DTT, 4 μM flavin adenine dinucleotide, 2 mM l-arginine, 2 mM NADPH (all from Sigma-Aldrich) and 4 μM 6R-5,6,7,8-tetrahydrobiopterin (Schircks Laboratoirs, Jona, Switzerland) (final volume 100 μl) for 120 min at 37°C. Then, 20 U/ml lactate dehydrogenase was added to stop the reaction by oxidation of residual NADPH. Product nitrite was measured by the Griess reaction. In addition, an extra well per sample was incubated with all reagents except l-arginine and NADPH to assess nonspecific absorbance, which was then subtracted from that of the test wells.
EMSA
Nuclear proteins were extracted as described by Schreiber et al. (33). For the preparation of radiolabeled probes representing standard consensus sequences of various transcription factors that are known to bind within the murine iNOS promoter, the following oligonucleotides were used: IFN regulatory factor 1 (IRF-1) sense 5′-GGAAGCGAAAATGAAATTG-3′, IRF-1 antisense 5′-TGAGTCAATTTCATTTTCG-3′; STAT-1 sense 5′-CATGTTATGCATATTCCTGTAAGTG-3′, STAT-1 antisense 5′-CGTGCACTTACAGGAATATGCATA-3′; NF-κB sense 5′-AGCTTCAGAGGGGACTTTCCGAGAGG-3′, NF-κB antisense 5′-TCGACCTCTCGGAAAGTCCCCTCTGA-3′; C/EBP-β sense 5′-AAGCTGCAGATTGCGCAATCTGCA-3′, C/EBP-β antisense 5′-CGTGCAGATTGCGCAATCTGCA-3′. For the preparation of double-stranded probes, oligomers were annealed, and overhanging ends were filled with [α-32P]dCTP (Amersham, Aylesbury, U.K.) and the three other nonradiolabeled dNTPs (Pharmacia, Piscataway, NJ) using Klenow enzyme (Amersham). For the preparation of unlabeled competitors, dCTP instead of [α-32P]dCTP was used. A total of 10 μg of nuclear extracts were preincubated with 2 μg of double-stranded poly(dI-dC)·poly(dI-dC) (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 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. After addition of 87% glycerol as a loading buffer, samples were separated on a 6% nondenaturing polyacrylamide gel, which was subsequently dried and exposed for autoradiography (34).
Transfections
iNOS promoter luciferase constructs with either the wild-type murine iNOS promoter or the iNOS promoter bearing point mutations at the consensus sequences of several transcription factors were generated as described previously (21). Transient transfections of RAW macrophages with these plasmids were performed by a lipofection method using the Cytofectene Transfection Reagent kit (Bio-Rad, Hercules, CA) according to the manufacturer’s protocol. Briefly, 2.5 × 105 cells in 2 ml of medium were seeded into 6-well plates and grown for 12 h. Then, 8 μl of Cytofectene transfection reagent per well were incubated with 2 μg of the iNOS-promoter plasmid and 0.25 μg of pRL-null (Promega, Mannheim, Germany) in 100 μl of serum-free medium for 20 min. Subsequently, the medium of the cells was removed and replaced by the cytofectene/plasmid mix filled to 1 ml per well with serum-containing medium. After 16 h, cells were washed with fresh medium and stimulated with IFN-γ (10 U/ml) and LPS (1 ng/ml) for 6–24 h before luciferase activity was determined by the dual luciferase system from Promega according to the manufacturer’s instructions. Firefly luciferase activity was corrected upon cotransfection of cells with a Renilla luciferase vector pRL-null (Promega).
Statistics
Differences in iNOS activity between the two cell lines were assessed by Student’s t test.
Results
To see whether different nitrite levels in RAW-21 and RAW-37 cells may be related to changes in iNOS activity, we stimulated cells with IFN-γ/LPS for 24 h and determined the enzymatic activity of iNOS. As can be seen from Fig. 1⇓a, iNOS activity was significantly higher in cytokine-induced Nramp1-bearing cells (RAW-37) compared with RAW-21. Moreover, the addition of iron decreased iNOS protein expression in both cell lines, whereas additional treatment of cytokine-activated RAW cells with the iron chelator desferrioxamine further increased enzymatic NO formation.
Effect of Nramp1 functionality on NO production in activated and S. typhimurium-infected RAW cells. a, iNOS enzyme activity assays were conducted as described in Materials and Methods. Cells were stimulated with IFN-γ (50 U/ml) and LPS (10 ng/ml) plus FeCl3 (50 μM) or desferrioxamine (100 μM). Data are mean ± SD of three independent experiments. b, Macrophages were infected with S. typhimurium and left untreated or stimulated with IFN-γ (100 U/ml) as detailed in Materials and Methods. NO formation was estimated by determination of nitrite concentration in tissue culture supernatants. Data are mean ± SD of three independent experiments. ∗, p < 0.01; and #, p < 0.05 when comparing RAW-21 with RAW-37 under the same treatment conditions as estimated by Student’s t test.
To see whether or not these changes may be related to alterations of RAW-21 or RAW-37 cells linked to the stable transfection and selection procedure, we performed transient transfections of parent RAW264.7 cells using plasmids LK-2 (Nramp1 antisense) and LK-3 (Nramp1 sense) using electroporation followed by incubation of cell for another 24 h and subsequent stimulation with IFN-γ (50 U/ml) and LPS (10 ng/ml) for another 20 h. As observed with stable transfectants, iNOS activity was significantly higher in Nramp1-expressing RAW cells compared with cells transiently or stably transfected with an Nramp1 antisense construct (results not shown). This suggested that the observed differences in iNOS expression between RAW-21 and RAW-37 cannot be related to alterations or mutations of cells as a function of the stable transfection and selection processes.
To explore the reliability of the results shown in Fig. 1⇑a in an infection model, we performed experiments in which RAW cells were infected with S. typhimurium. Infection of RAW-37 cells with these pathogens resulted in significantly increased NO formation compared with RAW-21 cells. Moreover, additional stimulation of infected macrophages with IFN-γ further increased NO formation in macrophages, which again was significantly higher in Nramp1-expressing cells (RAW-37) compared with RAW-21 (Fig. 1⇑b).
As a next step, we performed Northern blot analysis with mRNA obtained from RAW-21 and 37 cells stimulated with increasing dosages of IFN-γ, LPS, or both for 24 h. In RAW-21 cells the mRNA expression of iNOS following IFN-γ or IFN-γ/LPS combined treatment was significantly lower than in Nramp1-expressing RAW-37 macrophages, whereas the differences were not so prominent when LPS was used alone, although iNOS mRNA expression with the latter stimulus was rather low (Fig. 2⇓).
Effect of Nramp1 functionality on iNOS expression. Northern blots were done as described in Materials and Methods. iNOS mRNA-levels after stimulation of cells with the respective dosages of IFN-γ (units per milliliter) and/or LPS (nanograms per milliliter) are demonstrated. One of three representative experiments is shown.
To get more insights into the mechanism underlying different iNOS mRNA expression between RAW-21 and RAW-37 cells, we next performed time-course experiments. RAW-21 and RAW-37 macrophages were stimulated with IFN-γ (50 U/ml) and LPS (10 ng/ml) for 3–24 h, and iNOS mRNA levels were visualized by Northern blotting. Interestingly, both cell lines showed no differences during early iNOS induction. However, increased amounts of iNOS-mRNA were observed after prolonged cytokine stimulation for >6 h in RAW-37 cells that carry functional Nramp1 (Fig. 3⇓), whereas in RAW-21 cells, iNOS mRNA levels decreased over time.
Time-dependent changes in iNOS mRNA expression in activated RAW-21 and RAW-37 cells. iNOS mRNA levels after stimulation of cells with IFN-γ (50 U/ml) and LPS (10 ng/ml) for 3–24 h. Results from one of five similar experiments are shown.
Thus, the decline in iNOS mRNA levels in RAW-21 cells could be caused by diminished stabilization of iNOS mRNA or an altered transcription.
Therefore, we studied the effect of Nramp1 functionality on iNOS mRNA half life following stimulation of RAW-21 and 37 cells with IFN-γ/LPS and addition of actinomycin D, an inhibitor of nuclear transcription. As indicated by Fig. 4⇓, mRNA half life was not significantly different between Nramp1-expressing cells (RAW-37) and macrophages harboring the antisense Nramp1 (RAW-21).
Nramp1 functionality does not affect iNOS mRNA half life. Cells were stimulated with IFN-γ (50 U/ml) and LPS (10 ng/ml) for 24 h, and iNOS mRNA-levels were determined following the addition of actinomycin D (5 μg/ml) for 4–12 h before harvesting. One of four representative experiments is shown. The duration of autoradiography was adapted to get similar exposures in RAW-21 (48 h) and RAW-37 (16 h) cells.
To find out whether the influence of Nramp1 on NO formation is based on enhanced transcription of iNOS, we examined iNOS promoter activity using luciferase reporter gene assays. When cells were transiently transfected with full-length wild-type iNOS promoter luciferase constructs (Fig. 5⇓) and consecutively stimulated with IFN-γ and LPS for 0–24 h, relative luciferase activity (compared with a cotransfected Renilla construct used as an internal control) was significantly higher in RAW-37 cells than in RAW-21. This indicates that Nramp1 function leads to an increased transcriptional activity of the iNOS promoter.
Role of Nramp1 functionality for iNOS promoter activity. Cells were transiently transfected with wild-type iNOS promoter luciferase constructs as described in Materials and Methods. Relative luciferase activity after 0–24 h of stimulation with IFN-γ (10 U/ml) and LPS (1 ng/ml) is shown. Data are mean ± SD of four independent experiments. ∗, p < 0.05 when comparing RAW-21 with RAW-37 under the same treatment conditions as estimated by Student’s t test. #, p < 0.05, in comparison to iNOS promoter activity in unstimulated cells (0 h) in the respective treatment group.
Because the altered transfection may relate to different activity of transcription factors targeting the iNOS promoter, we next investigated the binding affinity of transcription factors that have previously been shown to be centrally involved in cytokine-mediated iNOS expression. Therefore, we performed EMSA with the corresponding radiolabeled consensus sequences of these transcription factors. As shown in Fig. 6⇓, we found increased binding activity of IRF-1 and STAT-1 over time in cellular extracts prepared from RAW-37 cells carrying functional Nramp1 compared with RAW-21 cells. In contrast, we found no clear differences in C/EBP-β (also termed NF-IL6) binding activity and NF-κB activity between RAW-37 and RAW-21 cells (Fig. 6⇓).
Modulation of transcription factor binding affinity in RAW-21 and RAW-37 cells. EMSA showing binding activity of several transcription factors in nuclear extracts derived from cells stimulated with IFN-γ (50 U/ml) and LPS (10 ng/ml) for 1–24 h. The specificity of binding was confirmed by cold competition with a 30-fold excess of the same unlabeled oligonucleotide (far right lane). One of four representative experiments is shown.
To find out whether the enhanced binding affinity of IRF-1 is also reflected by elevated amounts of IRF-1 mRNA, we performed Northern blots for murine IRF-1. Although IRF-1 mRNA expression was similarly induced in RAW-21 and RAW-37 cells after 3 h of cytokine stimulation, IRF-1 mRNA levels then declined in RAW-21 cells, whereas in RAW-37, cytoplasmic IRF-1 mRNA levels remained high (Fig. 7⇓). To see whether increased IRF-1 mRNA expression in Nramp1-bearing cells may account for enhanced iNOS transcription, we performed transient transfections with iNOS promoter luciferase constructs carrying a site-specific mutation at the binding site of IRF-1 within the iNOS promoter. Strikingly, the increase in relative luciferase activity that can be seen after cytokine stimulation of cells transfected with the wild-type iNOS promoter is abolished when the transfection is conducted with an iNOS promoter luciferase construct bearing a mutation at the IRF-1 binding site (Fig. 8⇓).
Nramp1 affects IRF-1 mRNA expression. Cytoplasmic IRF-1 mRNA-levels after stimulation of RAW-21 and RAW-37 cells with IFN-γ (50 U/ml) and LPS (10 ng/ml) for 3–24 h are plotted. One of three representative experiments is shown.
Mutation of the IRF-1 binding site within the iNOS promoter abolishes Nramp1-mediated regulation of iNOS expression. Transient transfection of cells with iNOS promoter luciferase constructs carrying a site-specific mutation at the IRF-1 binding site. Relative luciferase activity after 12 h of stimulation with IFN-γ (10 U/ml) and LPS (1 ng/ml) is shown. Data are mean ± SD of three independent experiments. ∗, p < 0.05 when comparing RAW-21 with RAW-37 under the same treatment conditions as estimated by Student’s t test. #, p < 0.05, in comparison to iNOS promoter activity in unstimulated cells (control) in the respective treatment group.
Discussion
In this study, we investigated the mechanism by which Nramp1 functionality increases NO formation. We found that Nramp1-expressing RAW264.7 cells present with increased NO formation and iNOS activity following cytokine stimulation, which can be related to enhanced and prolonged iNOS mRNA expression. In contrast to previous reports (28), at least from our data, this cannot be related to differences in iNOS mRNA half life.
However, the original data on mRNA half life of several IFN-dependent genes was generated with primary macrophages derived from different mouse strains (28, 35). In contrast, we used one murine macrophage-like cell line stably transfected with either functional (RAW-37) or nonfunctional (antisense) Nramp1 (RAW-21) (30), which should—apart from effects exerted by Nramp1 function—basically behave the same way, and thus avoid biasing of results by interstrain differences.
We propose that the elevated iNOS mRNA expression in Nramp1-carrying cells can be related to increased transcriptional activity. This is underscored by our experiments using iNOS promoter luciferase constructs transiently transfected into RAW cells. Interestingly, iNOS mRNA expression and transcription was not significantly different between RAW-21 and 37 cells after 6 h of stimulation; however, while transcriptional activity then decreased in RAW-21 cells, it remained increased in RAW-37 cells. When studying the binding affinity of critical transcription factors to the iNOS promoter, we found IRF-1 binding affinity to be increased over time in Nramp1 (RAW-37) cells. This increased binding affinity could be related to increased IRF-1 mRNA expression in RAW-37 cells. Finally, the critical role of IRF-1 to cause prolonged iNOS transcription was ascertained by the finding that upon transfection of an iNOS promoter construct carrying a site-specific mutation at the binding site for IRF-1, the relative luciferase activity after stimulation of cells with IFN-γ/LPS is no more different between RAW-37 and RAW-21 cells.
Interestingly, parallel to IRF-1, STAT-1 binding activity is increased in Nramp1-bearing cells, not only after prolonged cytokine stimulation but also in unstimulated cells, reflecting an increased basal activity of STAT-1. It is known that STAT-1 can bind to the IRF-1 promoter, resulting in enhanced IRF-1 transcription (36). Thus, the observed IRF-1 activation and the elevated IRF-1 mRNA levels are suggested to be a result of enhanced STAT-1 activity. Because IRF-1 is the essential transcription factor for induction of iNOS by IFN-γ (37, 38), the up-regulation of iNOS in Nramp1-resistant cells is correlated to an increased response of these cells to stimulation with IFN-γ. The mechanisms by which Nramp1 function affects the activity of IRF and/or other transcription factors remain to be shown. Because Nramp1 functions as an H+/bivalent cation antiporter, possible explanations would be iron deprivation from the cytoplasm or acidification of intracellular pH, which can both affect the function of critical transcription factors within the iNOS promoter (39, 40). However, Nramp1-mediated iron deprivation from the cytoplasm, resulting in iNOS up-regulation, seems unlikely, because binding activity of C/EBP-β, which is the critical transcription factor for iron-mediated iNOS expression (21), does not substantially differ between RAW-21 and RAW-37 cells.
Taken together, our results show that cells carrying functional Nramp1 express higher amounts of iNOS and NO, which can be related to increased iNOS transcription caused by an elevated IRF-1 expression and STAT-1 activation. Being aware of the central role of NO for host resistance against invading microbes (22, 41) and the inability of Nramp1-resistant cells to exert anti-microbial effects after pharmacological blockage of NO formation (27), it is suggestive that part of the protective function of Nramp1 to control infections with intracellular pathogens such as Mycobacterium, Salmonella, or Leishmania species (27, 29) can be related to its stimulatory effect toward increased and prolonged NO formation.
Acknowledgments
We thank Sabine Engl for excellent technical assistance.
Footnotes
-
↵1 Address correspondence and reprint requests to: Dr. Günter Weiss, Department of Internal Medicine, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria. E-mail address: guenter.weiss{at}uibk.ac.at
-
↵2 Abbreviations used in this paper: Nramp1, natural-resistance associated macrophage protein 1; iNOS, inducible NO synthase; IRF-1, IFN regulatory factor 1; dCTP, d cytidine 5′-triphosphate.
- Received January 27, 2003.
- Accepted June 6, 2003.
- Copyright © 2003 by The American Association of Immunologists