Leishmania pifanoi Amastigotes Avoid Macrophage Production of Superoxide by Inducing Heme Degradation
http://www.100md.com
感染与免疫杂志 2005年第12期
Department of Microbiology and Cell Science, University of Florida, Gainesville, Florida
ABSTRACT
Whereas infections of macrophages by promastigote forms of Leishmania mexicana pifanoi induce the production of superoxide, infections by amastigotes barely induce superoxide production. Several approaches were employed to gain insight into the mechanism by which amastigotes avoid eliciting superoxide production. First, in experiments with nitroblue tetrazolium, we found that 25% of parasitophorous vacuoles (PVs) that harbor promastigotes are positive for the NADPH oxidase complex, in contrast to only 2% of PVs that harbor amastigotes. Second, confocal microscope analyses of infected cells labeled with antibodies to gp91phox revealed that this enzyme subunit is found in PVs that harbor amastigotes. Third, in immunoblots of subcellular fractions enriched with PVs from amastigote-infected cells and probed with antibodies to gp91phox, only the 65-kDa premature form of gp91phox was found. In contrast, subcellular fractions from macrophages that ingested zymosan particles contained both the 91- and 65-kDa forms of gp91phox. This suggested that only the immature form of gp91phox is recruited to PVs that harbor amastigotes. Given that gp91phox maturation is dependent on the availability of heme, we found that infections by Leishmania parasites induce an increase in heme oxygenase 1 (HO-1), the rate-limiting enzyme in heme degradation. Infections by amastigotes performed in the presence of metalloporphyrins, which are inhibitors of HO-1, resulted in superoxide production by infected macrophages. Taken together, we propose that Leishmania amastigotes avoid superoxide production by inducing an increase in heme degradation, which results in blockage of the maturation of gp91phox, which prevents assembly of the NADPH oxidase enzyme complex.
INTRODUCTION
Leishmania spp. are dimorphic parasites that cause a spectrum of clinical presentations, ranging from cutaneous lesions to infection of visceral organs, in immunocompetent hosts. In the Americas, cutaneous and diffuse cutaneous leishmaniasis are often caused by parasites of the Leishmania mexicana complex (Leishmania amazonensis and Leishmania pifanoi). Infections with these parasites exhibit a limited tendency to self-resolve, which suggests the expression of mechanisms to resist elimination by the host.
Several studies have assessed whether infection of macrophages with Leishmania results in the production of superoxide. Taken together, the conclusions of those studies paint a mixed picture. A majority of studies have shown that there is limited superoxide production when macrophages are incubated with L. donovani promastigotes (4, 10, 16). Studies with promastigotes of another Leishmania species, Leishmania major, have sometimes shown that infection with these parasites triggers superoxide production by macrophages (13, 21, 26). However, when purified metacyclic L. major promastigotes were employed, compared to unselected stationary stage parasites, metacyclic promastigotes elicited minimal superoxide production by macrophages (13). Studies with Leishmania chagasi and parasites of the L. mexicana complex have mostly shown that infection with the promastigote form of these Leishmania species triggers macrophage production of superoxide (18, 26). So the mixed picture of superoxide production by macrophages in response to Leishmania infections might be explained in part by the fact that different Leishmania species can elicit different responses from the same host (27). Also, there are apparent differences in the Leishmania-killing mechanisms elaborated by human and murine macrophages (18). For example, in human macrophages, reactive nitrogen species play a minor role in the control of leishmaniasis (18, 42).
The amastigote form of Leishmania is the parasite form that persists and replicates in the infected host beyond a few hours of promastigote infection. A few studies have observed that infection with amastigotes results in limited superoxide production (10, 18). In vivo studies employing mice with genetically engineered defects in reactive nitrogen production (iNOS–/–) or superoxide formation (phox–/–) assessed the relative roles of both of these antimicrobial responses in the Leishmania donovani control (3, 30). These studies found that there was an early but limited alteration in the course of leishmaniasis in the absence of a functional NADPH oxidase enzyme complex. This implies that production of reactive oxygen species in this murine model plays a limited or secondary role to reactive nitrogen intermediates in the control of Leishmania infections. Since Leishmania parasites have been found to be susceptible to reactive oxygen intermediates in vitro (19, 33), one likely explanation for why superoxide might play a limited role in the control of leishmaniasis is that infection with the amastigote form of the parasite results in only limited superoxide production. The goal of this study is to elucidate the mechanism by which amastigotes of the Leishmania mexicana complex suppress or avoid superoxide production by macrophages.
Superoxide is the product of the multisubunit NADPH oxidase enzyme complex. This complex contains the membrane-bound cytochrome b558, which is composed of at least two polypeptides (gp91phox and p22phox) and two nonidentical heme groups that are associated with gp91phox (15, 31). The gp91 subunit is synthesized as a 58-kDa polypeptide, and after limited glycosylation in the endoplasmic reticulum, becomes a 65-kDa molecule (43). Thereafter, it traffics through the trans-Golgi network, where it is additionally glycosylated, acquires heme, and emerges as a molecule of 91 kDa (15, 31, 43). This processing or maturation of gp91phox increases its affinity for the other membrane resident subunit, p22. Four additional components of the NADPH oxidase enzyme, p40, p47, p67, and Rac2 are mostly found in the cytosol and associate with the membrane-bound components upon activation (31). Assembly of this enzyme complex on the target membrane is essential for the local release of optimal amounts of superoxide.
In this study, we assess superoxide production by murine macrophages infected with L. mexicana pifanoi amastigotes. We present evidence that in cells infected with this parasite form, the NADPH oxidase enzyme complex does not assemble on vacuoles that harbor the parasite. This is most likely the result of defective maturation of gp91phox in infected cells. Since the maturation of gp91phox is dependent on the availability of heme, we show that Leishmania infection induces heme degradation.
MATERIALS AND METHODS
Materials. Dulbecco's minimal essential medium and RPMI 1640 with L-glutamine were purchased from Mediatech, Inc. (Herndon, VA). Nitroblue tetrazolium (NBT) was purchased from Fischer Scientific (Fair Lawn, NJ). Tetrazolium WST-1 was obtained from Dojindo Laboratories (Kumamoto, Japan). Superoxide dismutase (SOD) was purchased from Sigma Chemicals (St. Louis, MO). Sn(IV) mesoporphyrin IX dichloride and cobalt protoporphyrin were obtained from FrontierScientific (Logan, Utah). Antibodies reactive to gp91phox (54.1 and NL7) were kind gifts of A. J. Jesaitis and J. Burritt (Montana State University, Bozeman, Montana). A rabbit polyclonal antibody to p22 (sc-20781) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal rabbit antibodies reactive to heme oxygenase 1 (HO-1) were obtained from Stressgen Biotechnologies (Victoria, BC, Canada). Antiserum to actin from clone JLA20 was obtained from the Iowa Developmental Studies Hybridoma Bank (Iowa City, IA).
Parasites. Leishmania mexicana pifanoi (MHOM/VE/60/Ltrod) amastigotes were maintained at 31°C in F-29 medium containing 20% HFBS (GIBCO BRL, Grand Island, New York) as previously reported (34). Promastigotes of this parasite line were grown at 23°C in complete Schneider's Drosophila melanogaster medium supplemented with 20% HIFBS and 10 μg/ml gentamicin. Some experiments have been repeated with L. mexicana pifanoi (MHOM/VE/57/LL1) obtained from the American Type Culture Collection (ATCC). Promastigotes of this line were cultured in complete Schneider's Drosophila medium. The amastigote forms were transformed and cultured in RPMI 1640-morpholineethanesulfonic acid (MES) (pH 5.5) medium supplemented with 20% fetal calf serum following protocols described by Debrabant et al. (14). Successful transformation of promastigotes to amastigotes was monitored by testing for the loss of reactivity to the gp46/M2 monoclonal antibody and expression of the 34-kDa form of P8 (34) in Western blots. Although data are not shown, Leishmania amazonensis (MHOM/BR/77/LTB0016) promastigotes were maintained in complete medium (Schneider's Drosophila medium; GIBCO BRL, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (GIBCO BRL, Grand Island, NY) and 10 μg/ml gentamicin at 23°C. The infectivity of parasites was maintained by periodic passage through BALB/c mice as reported previously (34). All parasites were used in the late stationary phase.
Macrophage. Immortalized RAW 264.7 murine macrophages were obtained from the ATCC. They were maintained at 37°C under a 5.5% CO2 atmosphere in RPMI 1640 complete medium supplemented with 10% heat-inactivated FBS, 100 μg/ml penicillin, 100 μg/ml streptomycin, and 10 μg/ml gentamicin.
WST-1 superoxide assay. Superoxide produced in culture was measured using a colorimetric microplate assay involving the superoxide-mediated reduction of the cell-permeable tetrazolium salt WST-1 to a water-soluble formazan precipitate (37). For the WST-1 superoxide assays, 96-well plates were seeded at 1.5 x 105 cells/well. The cultures were then incubated overnight (12 to 18 h) at 37°C at 5.5% CO2. Twenty minutes prior to infection, all macrophages were pretreated with 200 nM phorbol 12-myristate 13-acetate (PMA). WST-1 superoxide microplate assays were carried out in Hanks balanced salt solution and supplemented with 0.1% bovine serum albumin (BSA). The SOD inhibitible reaction was determined in parallel microplate wells containing 1.5 x 105 macrophages in a final volume of 0.1 ml Hanks balanced salt solution supplemented with 0.1% BSA. WST-1 (400 μM) with or without 100 U/ml SOD was added in conjunction with parasites at the time of infection. Samples were incubated at 34°C and 5% CO2 until they were taken out at the indicated intervals to be read at 438 nm using a Bio-Tek Power Wave 200 microplate reader. Superoxide production is expressed as the change in absorbance after subtracting the absorbance in SOD wells. In experiments where Sn(IV) mesoporphyrin (Sn MP) was used, the experiments were set up as above but the appropriate volume of Sn MP was added to each well to yield the desired concentrations. An identical plate of cells was prepared, to which Sn MP was added without WST-1. The absorbance values from this plate were subtracted from those obtained in the presence of WST-1. In experiments in which cobalt protoporphyrin IX (CoPP) was used, macrophages were preincubated for 2 h with a 30 μM concentration of this chemical. The macrophages were then incubated with WST-1 as discussed above. Here, too, an identical plate was set up with CoPP, and the experiment was performed without the addition of parasites. The absorbance values obtained here were subtracted from values obtained in the presence of parasites.
NBT assay. Macrophages were plated at 1 x 106 cells/dish in 60- by 15-mm petri dishes containing 12-mm round coverslips. Parasites were suspended in Dulbecco's minimal essential medium supplemented with FBS and containing 1 mg/ml NBT. The parasites were added to macrophages at a ratio of 5:1 parasites/cell. A discrete time of infection was achieved by incubating parasites with cells for 15 min at 34°C, 5% CO2 before washing the plates to remove unattached parasites and then reincubating the infected plates for the indicated time intervals. At the appropriate times, infection was terminated by the addition of 100% methanol to the culture for 5 min, followed by Giemsa (Sigma) staining of the cells on coverslips. Coverslips were mounted on glass slides with Gel Mount (Biomedia Corp., Foster City, CA). Cells were observed under differential interference contrast microscopy, and parasitized vacuoles were scored for the presence of the blue-black formazan precipitate, which forms upon the reduction of NBT by superoxide anion.
PV isolation. An enriched parasitophorous vacuole (PV) (ePV) fraction was obtained by following a previously described protocol (23). Briefly, at least 10 confluent dishes of RAW 264.7 macrophages in 100-mm petri dishes were incubated with parasites at a parasite/macrophage ratio of 4:1 and placed at 34°C and 5% CO2. After 30 min, free parasites were washed off with cold phosphate-buffered saline (PBS) and plates were returned to the incubator until the desired time. After 2 h of infection, plates were washed with cold PBS and cells were scraped into lysis buffer (20 mM HEPES, 0.5 mM EGTA, 0.25 M sucrose, and 0.1% gelatin) containing protease inhibitors (protease inhibitor cocktail; Roche, Mannheim, Germany). The cell suspension (1 ml per 10 plates) was lysed, and postnuclear supernatant (PNS) was recovered after low-speed centrifugation. The PNS was then loaded onto a sucrose step gradient (4 ml/step; 20%, 40%, 60% sucrose in gradient buffer [30 mM HEPES, 100 mM NaCl, 0.5 mM CaCl2, 0.5 mM MgCl2, pH 7.0]) and centrifuged at 700 x g for 25 min at 4°C. An ePV fraction was recovered from the 40/60% sucrose interface. The ePV fraction was then centrifuged at 12,000 x g. The protein and -hexosaminidase contents of this fraction were determined by the Bradford method.
The ePV fraction is composed of late endocytic vesicles and PVs. For affinity purification of PVs, the ePV fraction was incubated with polyclonal goat antiserum to the C-terminal tail of calnexin (Santa Cruz Biotechnology). This mixture was allowed to incubate for at least 1 h before the addition of protein G or donkey anti-goat antibody, conjugated with magnetic particles (Spherotech, Libertyville, Illinois). Antibody-bound vacuoles were positively selected with a magnet to yield the purified PV (pPV) fraction, and the supernatant fluid (V fraction) was saved for analysis. There is an enrichment of parasite molecules in the pPV; however, parasite molecules, most likely from broken parasites, are found in the remnant V fraction as well.
Western blot analysis. Equivalent protein aliquots (50 μg) of the PNS, ePV, pPV, and V fractions were separated on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted to nitrocellulose membranes according to established protocols. Magnetic particles in the pPV were removed by brief centrifugation after samples were boiled, before being loaded onto gels. For evaluation of HO-1 levels, 35 μg of lysates from infected cells, obtained at the times indicated, was resolved on 15% SDS-PAGE and then blotted onto nitrocellulose membranes. Membranes were blocked in Tris-buffered saline supplemented with 5% milk in Tris-buffered saline before incubation overnight in primary antibodies. After removal of primary antibodies and washing, membranes were incubated in the appropriate secondary antibodies conjugated to horseradish peroxidase. Washed blots were incubated with chemiluminescence (ECL kit; Amersham) reagents. Antibody reactivity was visualized by exposure of blots to X-ray film. Some blots were stripped by incubation at 50°C for 30 min in 20 mM 2-mercaptoethanol, 2% SDS, and 62.5 mM Tris-HCl, pH 6.8. They were then reprobed with other primary antibodies and processed as described above.
Immunofluorescence assays. Parasites were resuspended in complete macrophage medium and added to macrophages previously seeded on glass coverslips, at a 10:1 parasite-to-macrophage ratio. Infections were allowed to proceed at 34°C under a 5% CO2 atmosphere before washing plates to remove unattached parasites. Infections were terminated by adding an equal volume of 4% paraformaldehyde in PBS to the cultures. Coverslips were first incubated in 50 mM NH4Cl at room temperature for 15 min. Then coverslips were washed and incubated for 20 min in binding buffer (2% [wt/vol] BSA in 1x PBS) before incubation with primary antibodies. Coverslips were washed and then incubated with Alexa Fluor secondary antibodies reactive with mouse or rat immunoglobulin G (Molecular Probes, Eugene, OR). The nucleic acid dye 4',6'-diamidino-2-phenylindole dihydrochloride (DAPI; Molecular Probes) was added to the secondary antibody. Coverslips were washed and mounted on glass slides with Gel Mount (Biomedia Corp., Foster City, CA). Cells on coverslips that received secondary antibodies alone were always included in the analysis. Labeled cells were visualized and images captured on a laser scanning microscopy Pascal confocal microscope (Carl Zeiss, Inc.).
Statistics. SigmaPlot software was used to plot and analyze the data (Point Richmond, CA). Data are expressed as means ± standard errors (SE). t tests were used to analyze differences between groups. We defined statistical significance as a P value of <0.05.
RESULTS
Superoxide production by macrophages in response to L. pifanoi amastigotes. We elected to use Leishmania mexicana pifanoi, a member of the L. mexicana complex, in these studies because of available protocols to maintain axenic cultures of both the promastigote and amastigote forms of the parasite (14, 34). The capacity of these parasite forms to elicit superoxide production by macrophages was ascertained in WST-1 assays (37). Figure 1 shows that when RAW 264.7 macrophages are incubated with amastigotes, the macrophages produce negligible to low levels of superoxide. Incubation of macrophages with parasites at a ratio of 1:20 resulted in infection of greater than 50% of the macrophages. Other ratios of amastigotes to macrophages (5:1, 10:1, and 40:1) as well as antibody- and serum-opsonized amastigotes yielded similar results to those shown in Fig. 1. In contrast, significantly higher levels of superoxide were produced in parallel cultures of macrophages exposed to late-stationary-stage L. mexicana pifanoi promastigotes at 10:1 (P = 0.0004) and 20:1 parasite/macrophage ratios(P = 0.01). The response of macrophages to zymosan ingestion was also determined. Both nonopsonized zymosan (not shown) and antibody-opsonized zymosan elicited the release of superoxide as shown in Fig. 1. These results show that, whereas infection with the amastigote form of L. mexicana pifanoi barely induces the release of superoxide by macrophages, macrophages infected with the promastigote form produce significant levels of superoxide. Similar experiments have been performed with promastigotes of L. amazonensis, and we have found that they too elicit comparable responses to promastigotes of L. mexicana pifanoi. Also, infection of thioglycolate-elicited peritoneal cells from BALB/c mice yielded similar results (not shown).
The NADPH oxidase enzyme complex doesn't assemble on PVs that harbor amastigotes within infected cells. As an additional assessment of superoxide production, we determined whether the NADPH oxidase enzyme complex assembles around PVs that harbor Leishmania parasites. For these studies, we performed a widely accepted functional assay for the presence of the NADPH oxidase enzyme complex on a target membrane by assessing formazan precipitation as a result of the reduction of NBT by oxygen radicals produced by the enzyme. Representative images of NBT-positive parasitized vacuoles that harbor promastigotes and NBT-negative vacuoles that harbor amastigotes are shown in Fig. 2A and B, respectively. At 30 min postinfection, over 25% of PVs that harbor promastigotes were found to be NBT positive (Fig. 2C). There was a gradual decrease in the number of NBT-positive PVs through to the 6-h sampling point. In parallel experiments, less than 2% of vacuoles that harbored amastigotes could be scored as NBT positive. Pretreatment of macrophages with gamma interferon didn't augment the number of NBT-positive, amastigote-containing PVs (not shown).
Although most amastigote PVs were found to be NBT negative, those experiments didn't indicate at which point the assembly of the NADPH oxidase complex onto PVs is inhibited. Since formation of the NADPH oxidase complex can be initiated first by the recruitment of the membrane-bound subunits (15), we sought to determine whether the membrane subunits of the NADPH oxidase complex, particularly gp91phox, localize to amastigote PVs. Immunofluorescence staining of infected cells was performed to evaluate the distribution of gp91phox. We took advantage of the observation that after 1 h of uptake of zymosan particles or Leishmania parasites (promastigotes or amastigotes), over 95% reside in vacuoles that are Lamp1 positive. Lamp1 immunolabeling of vacuoles can therefore be used to delineate the contours of PVs. Figure 3 shows representative confocal images of macrophages infected with amastigotes (Fig. 3A, B, C, and D) or macrophages that engulfed zymosan particles (Fig. 3E, F, G, and H). Lamp1 and gp91phox staining are shown, and in the merged images, colocalization of gp91phox and Lamp1 is shown. As Fig. 3H shows, gp91phox is abundantly recruited to zymosan-containing vacuoles. In macrophages infected with amastigotes (Fig. 3D), there is a lot less gp91phox recruited to the membrane of amastigote-containing vacuoles than to zymosan-containing vacuoles. There are, however, patches of gp91phox on the PVs that harbor amastigotes. The pattern of gp91phox recruitment on promastigote-containing vacuoles is intermediate between the observations with zymosan and amastigotes (not shown). Taken together, these results confirm that the NADPH enzyme complex is minimally recruited to vacuoles that harbor Leishmania amastigotes, even though these vacuoles are sometimes reactive with antibodies to gp91phox
Full-length gp91 is absent from amastigote-infected parasitophorous vacuole fractions. Microscopic studies provide a powerful visual assessment of the distribution of molecules in cells, but they are less suitable for providing information about the structure of the molecule that is being monitored. We therefore used an additional approach to assess the recruitment of gp91phox to PVs that harbor Leishmania amastigotes. For these studies, we took advantage of an established protocol for isolating enriched PV fractions (8) and a more recently described protocol for obtaining affinity-purified PVs (23). Macrophages that had either ingested zymosan particles or been infected with amastigotes for 2 h were lysed, and subcellular fractions were obtained. Equivalent protein amounts of the PNS, ePV, pPV, and remaining V fractions were analyzed. The presence of gp91phox in these preparations was probed in Western blots using the 54.1 monoclonal antibody (5). The representative results in Fig. 4 show that the 91-kDa molecule which is reactive to monoclonal antibody 54.1, is present in the PNS, ePV, and pPV fractions obtained from macrophages that ingested zymosan particles. In contrast, gp91phox is not found in the ePV and pPV fractions obtained from amastigote-infected macrophages, even though gp91phox is detected in the PNS obtained after amastigote infections. Clearly evident though, in these fractions, is the 65-kDa endoplasmic form of gp91phox. This 65-kDa molecule was also detected with the NL7 monoclonal antibody (not shown), which is also specific for gp91phox (6).
We next probed for the presence of p22phox, the other membrane-bound subunit of the NADPH oxidase enzyme complex. Interactions between p22phox and gp91phox are essential for the stability of p22phox on intracellular membranes (15, 31). Figure 4 shows that p22phox is present in the subcellular fractions obtained from macrophages that had ingested zymosan particles. Although the PNS from amastigote-infected cells contains p22phox, much less is found in the ePV fraction and the rest of the vacuolar fractions obtained from these infected cells. This is most likely because of the absence of full-length gp91phox. We had previously found that Lamp1 is distributed throughout these subcellular fractions and can therefore be used as an internal control for protein loading (23). Taken together, these results show that, in amastigote-infected cells, the maturation of the p65 form of gp91phox to the full-length molecule is blocked. Reactivity of the 65-kDa form with antibodies to gp91phox might account for gp91phox-reactive PVs observed in confocal microscopy studies (Fig. 3).
Leishmania infection triggers an increase in HO-1 levels. The membrane-bound subunit of the NADPH oxidase enzyme, cytochrome b558, is composed of p22phox and gp91phox bound to heme. A couple of reports have shown that modulation of heme levels by either blocking their synthesis or inducing their degradation results in a decrease in gp91phox levels without affecting the abundance of p65 (15, 36). Changes in HO-1, the rate-limiting enzyme in heme degradation, was found to be a good indicator of heme degradation (36). Since gp91phox was not detected in subcellular fractions obtained from macrophages infected with amastigotes but the levels of p65 appeared to be unaffected, we performed experiments to determine whether infection of macrophages with these parasites induces heme degradation. On immunoblots, we assessed changes in the level of HO-1 after infection with Leishmania parasites or ingestion of zymosan particles. Figure 5A shows that infection with Leishmania amastigotes results in a significant increase in HO-1 levels as early as 1 h after infection and is sustained several hours postinfection. A plot of the densitometric scan from three identical experiments (Fig. 5B) confirms that the levels of HO-1 through the first 9 h postinfection are always significantly higher than the levels in resting cells (RawR). Curiously, infection with the promastigote form also results in an increase in HO-1 levels. It is noteworthy, though, that the level of HO-1 in cells exposed to amastigotes was always higher at 1 h after infection that in promastigote-infected cells. An increase in HO-1 levels in response to both parasite forms might explain why macrophages infected with promastigotes are progressively refractory to the production of superoxide (40). Ingestion of Zymosan particles resulted in minimal change in HO-1 levels after 1 h or 9 h (not shown). The level of actin in these samples was used as an internal control for sample loading (Fig. 5B).
Agents that modulate heme levels or heme activity alter superoxide production by macrophages infected with amastigotes or promastigotes. In light of the observation that infection induces HO-1, two experimental designs were employed to determine whether HO-1 plays a role in superoxide production by macrophages infected with Leishmania parasites. There are several synthetic heme analogs that can regulate heme availability either by inducing its degradation through the induction of HO-1 (CoPP) or by blocking its degradation by competitive inhibition of HO-1 (22, 24, 25). In the first design, infections were performed in the presence of tin mesoporphyrin (Sn MP), which is an inhibitor of HO-1 activity, to determine whether blocking of HO-1 activity would relieve macrophages to produce superoxide when they are exposed to amastigotes. In the second approach, macrophages were pretreated with CoPP, which induces HO-1 production. These experiments were designed to assess what the effect of higher levels of HO-1 at the time of infection would be on superoxide production by macrophages in response to Leishmania infection. Figures 6 and 7 show the results of WST-1 assays with these compounds, which were compiled from three experiments. Macrophages infected with amastigotes in the presence of 50 μM Sn MP produced significantly higher levels of superoxide than macrophages that were incubated with amastigotes alone (P = 0.0095) (Fig. 6A). Infections of macrophages with promastigotes in the presence of Sn MP did not have a significant effect (P = 0.321) on superoxide production by these macrophages (Fig. 6B).
Complementary results were obtained when macrophages were preincubated with CoPP. Figure 7 shows that superoxide production by macrophages that were preincubated for 2 h with CoPP and then incubated with promastigotes produced significantly less superoxide (P < 0.05) than macrophages that were incubated with promastigotes without preincubation with CoPP. Preincubation with CoPP also affected superoxide production by amastigote-infected macrophages: it further reduced the minimal superoxide levels that were produced in response to amastigotes. These experiments show a significant dependence of the capacity to produce superoxide on HO-1 levels. We should note that a new batch of WST-1 reagent was used in these latter experiments, which resulted in higher absorbance readings. Nonetheless, the finding that amastigote-infected macrophages produced significantly less superoxide than promastigote-infected macrophages was consistent.
Amastigotes suppress superoxide production by macrophages in coculture experiments. We wondered what effect amastigotes would have on particulate stimuli that can trigger macrophage superoxide production. To address this question, coincubation studies with amastigotes and promastigotes or amastigotes and zymosan were performed in the presence of PMA. We presumed that zymosan particles and promastigotes engage different macrophage receptors to activate superoxide production. Macrophages were incubated with promastigotes, and then in parallel cultures, increasing numbers of live or fixed amastigotes were added to the promastigote infection. WST-1 assays that assessed superoxide production showed that when the increasing number of amastigotes was used, there was a reduction of superoxide production by macrophages in response to promastigote infection (Fig. 8a). In the presence of a 30:1 amastigote-to-macrophage ratio, promastigote-infected macrophages released 40% less superoxide (P < 0.03) than in infections with promastigotes alone. To rule out the likelihood that amastigotes were merely outcompeting promastigotes for interactions with macrophages, heat-killed (not shown) or paraformaldehyde-fixed amastigotes were added to promastigote cultures in comparable numbers to the live parasite experiments. Figure 8b shows that paraformaldehyde-fixed parasites had no significant effect on promastigote-elicited superoxide production.
We also determined the effect of amastigotes on zymosan elicitation of superoxide production by macrophages. As Fig. 8C shows, coaddition of amastigotes at a ratio of 20:1 (parasite to macrophages) with zymosan (10:1) resulted in greater that 80% reduction in superoxide production by the macrophages. For comparison, promastigotes and zymosan particles were added together on macrophages, and superoxide production was assessed. Unlike the scenario with amastigotes, coaddition of zymosan particles with promastigotes on macrophages resulted in greater levels of superoxide production. These results show that Leishmania amastigotes can inhibit superoxide production elicited by different stimuli.
DISCUSSION
Reactive oxygen radicals are some of the potent antimicrobial responses that phagocytes elaborate to limit the growth of microorganisms. To survive and thrive, intracellular pathogens have to either avoid elicitation of oxygen radicals or resist their effects. In this report, we provide evidence that supports the finding that murine macrophages produce minimal levels of superoxide when they internalize the amastigote form of parasites of the L. mexicana complex. Three experimental approaches were employed to elucidate the underlying mechanism(s) that enables this phenomenon. First, in a histochemical assay using nitroblue tetrazolium, which is reduced to a formazan precipitate on membranes that contain the functional NADPH oxidase enzyme, we found that very few vacuoles that harbor the amastigote form of the parasite contain the NADPH oxidase complex (Fig. 2). This was contrasted with infections with the promastigote form, where 23% of those vacuoles were positive for the NADPH oxidase complex. The advantage of this assay is that it is minimally affected by the superoxide scavenger, super oxide dismutase, which is located in the mitochondria and the cytosol (12). Second, immunolocalization studies of gp91phox revealed that even though the NADPH oxidase enzyme complex isn't present on amastigote-containing vacuoles, some of these vacuoles are positive for gp91phox (Fig. 3). This raised the possibility that gp91phox molecules that are recruited to these vacuoles might not always be functional. Third, whereas subcellular fractions enriched with vacuoles that harbor zymosan particles contain detectable levels of gp91phox, comparable subcellular fractions from Leishmania amastigote-infected cells were devoid of the mature form of gp91phox, even though detectable levels of gp91phox were found in the postnuclear fraction. Instead, the precursor form of gp91phox, p65, is readily detected in these subcellular fractions. Taken together, these experiments suggested that amastigotes avoid the elicitation of superoxide by promoting activities that result in the blockage of the maturation of gp91phox and thus prevent the formation of functional enzyme complexes in infected cells.
Several microorganisms have been shown to inhibit the production of superoxide by macrophages and neutrophils. Some examples are Yersinia enterocolitica (41), Coxiella burnetii (1), Chlamydia trachomatis (38), and the agent of human granulocytic ehrlichiosis (HGE agent) (7, 28), just to name a few. These organisms suppress the release of superoxide when neutrophils are stimulated by known inducers of superoxide. From this group of organisms, the HGE agent was found to be somewhat unique because it suppresses superoxide release when neutrophils are exposed to stimuli that induce superoxide production through diverse mechanisms (29, 30). The capacity to inhibit superoxide release induced by multiple stimuli is an indication that the effect of this organism is at a point where several signaling pathways converge. In this regard, our observations with Leishmania amastigotes are most similar to the observations with the HGE agent. In Fig. 8, we showed that Leishmania amastigotes, too, suppress superoxide production by macrophages that were exposed to zymosan particles and promastigotes in the presence of PMA.
The regulation of superoxide production by microorganisms can occur at several points in their interaction. In this study, we focused on activities that affect the posttranslational processing of gp91phox, which eventually affects the assembly of the enzyme subunits. However, some studies, particularly those employing the HGE agent, have investigated the likelihood that suppression of superoxide production is a consequence of a targeted block in the transcription of gp91phox (2, 39). Those observations were made after a 5-day incubation of the HGE agent with neutrophils (2). Other studies in that infection system that have investigated the suppression of superoxide production in a much shorter frame (within 2 h) have found no changes in the level of gp91phox (29), even though the inhibition of superoxide production is significant by then. This suggests that modulating the transcription of gp91phox might not be a relevant strategy for the immediate survival of internalized organisms.
Gp91phox is synthesized as a 56-kDa molecule in the endoplasmic reticulum, where it undergoes limited glycosylation and transforms into a 65-kDa protein. Further processing of p65 occurs beyond the endoplasmic reticulum, where it is additionally glycosylated and acquires two heme molecules (31). Only the full-length gp91phox can stably associate with p22, the other membrane-bound subunit of the NADPH oxidase molecule (31). It was recently shown that induction of heme degradation with succinyl acetone, results in inhibition of gp91phox maturation into the fully glycosylated form, without affecting the levels of p65 (15). The addition of heme reversed the effect of succinyl acetone (43). In our studies, Western blot analysis of gp91phox distribution in subcellular fractions of amastigote-infected cells (Fig. 4) indicated that these parasites cause a comparable blockage of gp91phox maturation, as does succinyl acetone. This led us to hypothesize that infection with amastigotes results in an increase in HO-1, as does succinyl acetone, which results in impaired gp91phox maturation and limited superoxide production.
HO-1, also called HSP 32, is the rate-limiting enzyme in the degradation of heme (25). It catalyzes the conversion of heme to biliverdin. HO-1 is induced by a wide variety of stimuli. HO-1 levels can be rapidly induced, and maximal levels can be reached in 1 h (17). It is obvious from the studies presented here that both the promastigote and amastigote forms of the parasite induce macrophages to upregulate their levels of HO-1. However, it appears that amastigotes trigger a much quicker rise in HO-1 than is observed with promastigotes (Fig. 5). There is some evidence that promastigote and amastigote forms of Leishmania exhibit differences in heme metabolism (synthesis and degradation) (9, 35), but it is unlikely that such activities within the parasite contribute to the rapid effect on macrophage HO-1 levels.
Although we cannot rule out the likelihood that amastigotes might suppress superoxide production through alternate mechanisms, we are confident that the induction of HO-1 plays a significant role in the suppression of superoxide production. Two approaches were employed to fully implicate the induction of HO-1 in the underlying mechanism employed by amastigotes to avoid superoxide production. In the first, when HO-1 activity was blocked with competitive inhibitors, more superoxide is produced by macrophages in response to amastigote infection. In the second approach, induction of HO-1 prior to infection resulted in reduced superoxide production in response to infection. Metalloporphyrins, such as zinc protoporphyrin and Sn MP, are competitive inhibitors of HO-1 activity (22, 24) that have been used in the treatment of infants with neonatal jaundice, where the accumulation of bilirubin, a downstream metabolite of heme, is believed to be part of the pathophysiology of disease (20). We observed that Sn MP significantly reversed the inhibitory effect of amastigote infection on superoxide production (Fig. 6). Interestingly, Sn-MP didn't have a significant effect on superoxide production in promastigote infections. But preincubation with CoPP, which induces HO-1 production, resulted in a reduction of superoxide production in response to infection with promastigotes. This later result suggests that there might be an HO-1 threshold at which point its effect on heme levels has a significant effect on heme-containing compounds like gp91phox. So even though there is only a small difference in the HO-1 levels induced by promastigotes and amastigotes after 1 h (Fig. 5), that difference might be sufficient to result in a greater suppression of superoxide production by amastigotes.
The finding that Leishmania infections in the presence of metalloporphyrins result in increased superoxide production is an important observation. Since it has previously been shown that Leishmania parasites are susceptible to superoxide-mediated killing (19, 33), a strategy that augments superoxide in the vicinity of these parasites might result in the control of Leishmania infections. There is already extensive experience with using metalloporphyrins in humans, so it should be feasible to adapt them for use in this disease. Initial studies to assess their usefulness in controlling Leishmania infections are being considered in experimental models of the disease.
Amastigotes block superoxide production by other stimuli. Coincubation of amastigotes with promastigotes or zymosan particles muted the response of macrophages to these potent stimuli of superoxide production (Fig. 7). Fixed or heat-killed amastigotes had minimal effects on the macrophage response to promastigotes. This would suggest that live amastigotes actively alter processes within the macrophage that results in their refractoriness to otherwise potent stimuli. Since it is most probable that promastigotes, amastigotes, and zymosan particles engage different signaling pathways when they are engulfed by cells, the suppression of superoxide production of these stimuli suggest that the effect of amastigotes occurs at a point where all the signals converge.
Taken together, the studies presented here show that amastigotes not only avoid the elicitation of superoxide production during their internalization but also render the macrophages refractory to superoxide production by otherwise potent stimuli. This effect on heme metabolism is an evasion strategy that might be employed by other organisms.
Additionally, while the manuscript was being revised, a paper by Chauveau et al. (11) showed that induction of HO-1 by CoPP pretreatment resulted in a reduction in superoxide production by dendritic cells in response to lipopolysaccharide stimulation.
ACKNOWLEDGMENTS
We thank Algirdas J. Jesaitis and James Burritt at Montana State University for generous gifts of anti-gp91phox and p22 monoclonal antibodies (54.1, NL7, and 44.1). We also thank Mary Wilson for reading the manuscript and for suggestions on improving it.
This study was funded by grant AI048739 from the National Institutes of Health.
REFERENCES
1. Baca, O. G., M. J. Roman, R. H. Glew, R. F. Christner, J. E. Buhler, and A. S. Aragon. 1993. Acid phosphatase activity in Coxiella burnetii: a possible virulence factor. Infect. Immun. 61:4232-4239.
2. Banerjee, R., J. Anguita, D. Roos, and E. Fikrig. 2000. Cutting edge: infection by the agent of human granulocytic ehrlichiosis prevents the respiratory burst by down-regulating gp91phox. J. Immunol. 164:3946-3949.
3. Blos, M., U. Schleicher, F. J. Soares Rocha, U. Meissner, M. Rollinghoff, and C. Bogdan. 2003. Organ-specific and stage-dependent control of Leishmania major infection by inducible nitric oxide synthase and phagocyte NADPH oxidase. Eur. J. Immunol. 33:1224-1234.
4. Brandonisio, O., M. A. Panaro, R. Marzio, A. Marangi, S. M. Faliero, and E. Jirillo. 1994. Impairment of the human phagocyte oxidative responses caused by Leishmania lipophosphoglycan (LPG): in vitro studies. FEMS Immunol. Med. Microbiol. 8:57-62.
5. Burritt, J. B., M. T. Quinn, M. A. Jutila, C. W. Bond, and A. J. Jesaitis. 1995. Topological mapping of neutrophil cytochrome b epitopes with phage-display libraries. J. Biol. Chem. 270:16974-16980.
6. Burritt, J. B., T. R. Foubert, D. Baniulis, C. I. Lord, R. M. Taylor, J. S. Mills, T. D. Baughan, D. Roos, C. A. Parkos, and A. J. Jesaitis. 2003. Functional epitope on human neutrophil flavocytochrome b558. J. Immunol. 170:6082-6089.
7. Carlyon, J. A., D. Abdel-Latif, M. Pypaert, P. Lacy, and E. Fikrig. 2004. Anaplasma phagocytophilum utilizes multiple host evasion mechanisms to thwart NADPH oxidase-mediated killing during neutrophil infection. Infect. Immun. 72:4772-4783.
8. Chakraborty, P., S. Sturgill-Koszycki, D. G. Russell. 1994. Isolation and characterization of pathogen-containing phagosomes. Methods Cell Biol. 45:261-276.
9. Chang, C. S., and K. P. Chang. 1985. Heme requirement and acquisition by extracellular and intracellular stages of Leishmania mexicana amazonensis. Mol. Biochem. Parasitol. 16:267-276.
10. Channon, J. Y., M. B. Roberts, J. M. Blackwell. 1984. A study of the differential respiratory burst activity elicited by promastigotes and amastigotes of Leishmania donovani in murine resident peritoneal macrophages. Immunology 53:345-355.
11. Chauveau, C., S. Remy, P. J. Royer, M. Hill, S. Tanguy-Royer, F. X. Hubert, L. Tesson, R. Brion, G. Beriou, M. Gregoire, R. Josien, M. C. Cuturi, and I. Anegon. 2005. Heme oxygenase-1 expression inhibits dendritic cell maturation and proinflammatory function but conserves IL-10 expression. Blood 106:1694-1702. (First published 26 May 2005; doi:10.1182/blood-2005-02-0494.)
12. Crapo, J. D., T. Oury, C. Rabouille, J. W. Slot, and L.-Y. Chang. 1992. Copper, zinc superoxide dismutase is primarily a cytosolic protein in human cells. Proc. Natl. Acad. Sci. USA 89:10405-10409.
13. Da Silva, R. P., B. F. Hall, K. A. Joiner, and D. L. Sacks. 1989. CR1, the C3b receptor, mediates binding of infective Leishmania major metacyclic promastigotes to human macrophages. J. Immunol. 143:617-622.
14. Debrabant, A., M. B. Joshi, P. F. Pimenta, and D. M. Dwyer. 2004. Generation of Leishmania donovani axenic amastigotes: their growth and biological characteristics. Int. J. Parasitol. 34:205-217.
15. DeLeo, F. R., J. B. Burritt, L. Yu, A. J. Jesaitis, M. C. Dinauer, and W. M. Nauseef. 2000. Processing and maturation of flavocytochrome b558 include incorporation of heme as a prerequisite for heterodimer assembly. J. Biol. Chem. 275:13986-13993.
16. Descoteaux, A., and S. J. Turco. 1993. The lipophosphoglycan of Leishmania and macrophage protein kinase C. Parasitol. Today 9:468-471.
17. Ewing, J. F., S. N. Haber, and M. D. Maines. 1992. Normal and heat-induced patterns of expression of heme oxygenase-1 (HSP32) in rat brain: hyperthermia causes rapid induction of mRNA and protein. J. Neurochem. 58:4023-4029.
18. Gantt, K. R., T. L. Goldman, M. A. Miller, M. L. McCormick, S. M. B. Jeronimo, E. T. Nascimento, B. E. Britigan, and M. E. Wilson. 2001. Oxidative responses of human and murine macrophages during phagocytosis of Leishmania chagasi. J. Immunol. 167:893-901.
19. Haidaris, C. G., and P. F. Bonventre. 1982. A role for oxygen-dependent mechanisms in killing of Leishmania donovani tissue forms by activated macrophages. J. Immunol. 129:850-855.
20. Ip, S., M. Chung, J. Kulig, R. O'Brien, R. Sege, S. Glicken, M. J. Maisels, and J. Lau. 2004 An evidence-based review of important issues concerning neonatal hyperbilirubinemia. Pediatrics 114:e130-e153.
21. Kantengwa, S., I. Muller, J. Louis, and B. S. Polla. 1995. Infection of human and murine macrophages with Leishmania major is associated with early parasite heat shock protein synthesis but fails to induce a host cell stress response. Immunol. Cell Biol. 73:73-80.
22. Kappas, A., and G. S. Drummond. 1985. Metalloporphyrins: a class of compounds of pharmacological interests. BioEssays 3:256-259.
23. Kima, P. E., and W. Dunn. 2005. Exploiting calnexin expression on phagosomes to isolate Leishmania parasitophorous vacuoles. Microb. Pathog. 38:139-145.
24. Maines, M. D. 1981. Zinc-protoporphyrin is a selective inhibitor of heme oxygenase activity in the neonatal rat. Biochim. Biophys. Acta 673:339-350.
25. Maines, M. D. 1997. The heme oxygenase system: a regulator of second messenger gases. Annu. Rev. Pharmacol. Toxicol. 37:517-554.
26. Mallinson, D. J., and G. H. Coombs. 1989. Interaction of Leishmania metacyclics with macrophages. Int. J. Parasitol. 19:647-656.
27. McMahon-Pratt, D., and J. Alexander. 2004. Does the Leishmania major paradigm of pathogenesis and protection hold for New World cutaneous leishmaniasis or the visceral disease Immunol. Rev. 201:206-224.
28. Mott, J., and Y. Rikihisa. 2000. Human granulocytic ehrlichiosis agent inhibits superoxide anion generation by human neutrophils. Infect. Immun. 68:6697-6703.
29. Mott, J., Y. Rikihisa, and S. Tsunawaki. 2002. Effects of Anaplasma phagocytophila on NADPH oxidase components in human neutrophils and HL-60 cells. Infect. Immun. 70:1359-1366.
30. Murray, H. W., and C. F. Nathan. 1999. Macrophage microbicidal mechanisms in vivo: reactive nitrogen versus oxygen intermediates in the killing of intracellular visceral Leishmania donovani. J. Exp. Med. 189:741-746.
31. Nauseef, W. M. 2004. Assembly of the phagocyte NADPH oxidase. Histochem. Cell Biol. 122:277-291.
32. Passwell, J. H., R. Shor, J. Smolen, and C. L. Jaffe. 1994. Infection of human monocytes by Leishmania results in a defective oxidative burst. Int. J. Exp. Pathol. 75:277-284.
33. Reiner, N. E., and J. W. Kazura. 1982. Oxidant-mediated damage of Leishmania donovani promastigotes. Infect. Immun. 36:1023-1027.
34. Soong, L., S. M. Duboise, P. Kima, and D. McMahon-Pratt. 1995. Leishmania pifanoi amastigote antigens protect mice against cutaneous leishmaniasis. Infect. Immun. 63:3559-3566.
35. Srivastava, P., G. D. Sharma, K. K. Kamboj, A. K. Rastogi, and W. C. Pandey. 1997. Heme metabolism in promastigotes of Leishmania donovani. Mol. Cell. Biochem. 171:65-68.
36. Taille, C., J. El-Benna, S. Lanone, M. C. Dang, E. Ogier-Denis, M. Aubier, and J. Boczkowski. 2004. Induction of heme oxygenase-1 inhibits NAD(P)H oxidase activity by down-regulating cytochrome b558 expression via the reduction of heme availability. J. Biol. Chem. 279:28681-28688.
37. Tan, A. S., and M. V. Berridge. 2000. Superoxide produced by activated neutrophils efficiently reduces the tetrazolium salt, WST-1 to produce a soluble formazan: a simple colorimetric assay for measuring respiratory burst activation and for screening anti-inflammatory agents. J. Immunol. Methods 238:59-68.
38. Tauber, A. I., N. Pavlotsky, J. S. Lin, and P. A. Rice. 1989. Inhibition of human neutrophil NADPH oxidase by Chlamydia serovars E, K, and L2. Infect. Immun. 57:1108-1112.
39. Thomas, V., S. Samanta, C. Wu, N. Berliner, and E. Fikrig. 2005. Anaplasma phagocytophilum modulates gp91phox gene expression through altered interferon regulatory factor 1 and PU.1 levels and binding of CCAAT displacement protein. Infect. Immun. 73:208-218.
40. Uchmuller-Rouiller, Y., and J. Mauel. 1987. Impairment of the oxidative metabolism of mouse peritoneal macrophages by intracellular Leishmania spp. Infect. Immun. 55:587-593.
41. Visser, L. G., E. Seijmonsbergen, P. H. Nibbering, P. J. van den Broek, and R. van Furth. 1999. Yops of Yersinia enterocolitica inhibit receptor-dependent superoxide anion production by human granulocytes. Infect. Immun. 67:1245-1250.
42. Vouldoukis, I., P.-A. Becherel, V. Riveros-Moreno, M. Arock, O. da Silva, P. Debre, D. Mazier, and M. D. Mossalayi. 1997. Interleukin-10 and interleukin-4 inhibit intracellular killing of Leishmania infantum and Leishmania major by human macrophages by decreasing nitric oxide generation. Eur. J. Immunol. 27:860-865.
43. Yu, L., F. R. DeLeo, K. J. Biberstine-Kinkade, J. Renee, W. M. Nauseef, and M. C. Dinauer. 1999. Biosynthesis of flavocytochrome b558. gp91(phox) is synthesized as a 65-kDa precursor (p65) in the endoplasmic reticulum. J. Biol. Chem. 274:4364-4369.
44. Yu, L., L. Zhen, and M. C. Dinnauer. 1997. Biosynthesis of the phagocyte NADPH oxidase cytochrome b558. J. Biol. Chem. 272:27288-27294.(Nam-Kha Pham, Jennifer Mo)
ABSTRACT
Whereas infections of macrophages by promastigote forms of Leishmania mexicana pifanoi induce the production of superoxide, infections by amastigotes barely induce superoxide production. Several approaches were employed to gain insight into the mechanism by which amastigotes avoid eliciting superoxide production. First, in experiments with nitroblue tetrazolium, we found that 25% of parasitophorous vacuoles (PVs) that harbor promastigotes are positive for the NADPH oxidase complex, in contrast to only 2% of PVs that harbor amastigotes. Second, confocal microscope analyses of infected cells labeled with antibodies to gp91phox revealed that this enzyme subunit is found in PVs that harbor amastigotes. Third, in immunoblots of subcellular fractions enriched with PVs from amastigote-infected cells and probed with antibodies to gp91phox, only the 65-kDa premature form of gp91phox was found. In contrast, subcellular fractions from macrophages that ingested zymosan particles contained both the 91- and 65-kDa forms of gp91phox. This suggested that only the immature form of gp91phox is recruited to PVs that harbor amastigotes. Given that gp91phox maturation is dependent on the availability of heme, we found that infections by Leishmania parasites induce an increase in heme oxygenase 1 (HO-1), the rate-limiting enzyme in heme degradation. Infections by amastigotes performed in the presence of metalloporphyrins, which are inhibitors of HO-1, resulted in superoxide production by infected macrophages. Taken together, we propose that Leishmania amastigotes avoid superoxide production by inducing an increase in heme degradation, which results in blockage of the maturation of gp91phox, which prevents assembly of the NADPH oxidase enzyme complex.
INTRODUCTION
Leishmania spp. are dimorphic parasites that cause a spectrum of clinical presentations, ranging from cutaneous lesions to infection of visceral organs, in immunocompetent hosts. In the Americas, cutaneous and diffuse cutaneous leishmaniasis are often caused by parasites of the Leishmania mexicana complex (Leishmania amazonensis and Leishmania pifanoi). Infections with these parasites exhibit a limited tendency to self-resolve, which suggests the expression of mechanisms to resist elimination by the host.
Several studies have assessed whether infection of macrophages with Leishmania results in the production of superoxide. Taken together, the conclusions of those studies paint a mixed picture. A majority of studies have shown that there is limited superoxide production when macrophages are incubated with L. donovani promastigotes (4, 10, 16). Studies with promastigotes of another Leishmania species, Leishmania major, have sometimes shown that infection with these parasites triggers superoxide production by macrophages (13, 21, 26). However, when purified metacyclic L. major promastigotes were employed, compared to unselected stationary stage parasites, metacyclic promastigotes elicited minimal superoxide production by macrophages (13). Studies with Leishmania chagasi and parasites of the L. mexicana complex have mostly shown that infection with the promastigote form of these Leishmania species triggers macrophage production of superoxide (18, 26). So the mixed picture of superoxide production by macrophages in response to Leishmania infections might be explained in part by the fact that different Leishmania species can elicit different responses from the same host (27). Also, there are apparent differences in the Leishmania-killing mechanisms elaborated by human and murine macrophages (18). For example, in human macrophages, reactive nitrogen species play a minor role in the control of leishmaniasis (18, 42).
The amastigote form of Leishmania is the parasite form that persists and replicates in the infected host beyond a few hours of promastigote infection. A few studies have observed that infection with amastigotes results in limited superoxide production (10, 18). In vivo studies employing mice with genetically engineered defects in reactive nitrogen production (iNOS–/–) or superoxide formation (phox–/–) assessed the relative roles of both of these antimicrobial responses in the Leishmania donovani control (3, 30). These studies found that there was an early but limited alteration in the course of leishmaniasis in the absence of a functional NADPH oxidase enzyme complex. This implies that production of reactive oxygen species in this murine model plays a limited or secondary role to reactive nitrogen intermediates in the control of Leishmania infections. Since Leishmania parasites have been found to be susceptible to reactive oxygen intermediates in vitro (19, 33), one likely explanation for why superoxide might play a limited role in the control of leishmaniasis is that infection with the amastigote form of the parasite results in only limited superoxide production. The goal of this study is to elucidate the mechanism by which amastigotes of the Leishmania mexicana complex suppress or avoid superoxide production by macrophages.
Superoxide is the product of the multisubunit NADPH oxidase enzyme complex. This complex contains the membrane-bound cytochrome b558, which is composed of at least two polypeptides (gp91phox and p22phox) and two nonidentical heme groups that are associated with gp91phox (15, 31). The gp91 subunit is synthesized as a 58-kDa polypeptide, and after limited glycosylation in the endoplasmic reticulum, becomes a 65-kDa molecule (43). Thereafter, it traffics through the trans-Golgi network, where it is additionally glycosylated, acquires heme, and emerges as a molecule of 91 kDa (15, 31, 43). This processing or maturation of gp91phox increases its affinity for the other membrane resident subunit, p22. Four additional components of the NADPH oxidase enzyme, p40, p47, p67, and Rac2 are mostly found in the cytosol and associate with the membrane-bound components upon activation (31). Assembly of this enzyme complex on the target membrane is essential for the local release of optimal amounts of superoxide.
In this study, we assess superoxide production by murine macrophages infected with L. mexicana pifanoi amastigotes. We present evidence that in cells infected with this parasite form, the NADPH oxidase enzyme complex does not assemble on vacuoles that harbor the parasite. This is most likely the result of defective maturation of gp91phox in infected cells. Since the maturation of gp91phox is dependent on the availability of heme, we show that Leishmania infection induces heme degradation.
MATERIALS AND METHODS
Materials. Dulbecco's minimal essential medium and RPMI 1640 with L-glutamine were purchased from Mediatech, Inc. (Herndon, VA). Nitroblue tetrazolium (NBT) was purchased from Fischer Scientific (Fair Lawn, NJ). Tetrazolium WST-1 was obtained from Dojindo Laboratories (Kumamoto, Japan). Superoxide dismutase (SOD) was purchased from Sigma Chemicals (St. Louis, MO). Sn(IV) mesoporphyrin IX dichloride and cobalt protoporphyrin were obtained from FrontierScientific (Logan, Utah). Antibodies reactive to gp91phox (54.1 and NL7) were kind gifts of A. J. Jesaitis and J. Burritt (Montana State University, Bozeman, Montana). A rabbit polyclonal antibody to p22 (sc-20781) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal rabbit antibodies reactive to heme oxygenase 1 (HO-1) were obtained from Stressgen Biotechnologies (Victoria, BC, Canada). Antiserum to actin from clone JLA20 was obtained from the Iowa Developmental Studies Hybridoma Bank (Iowa City, IA).
Parasites. Leishmania mexicana pifanoi (MHOM/VE/60/Ltrod) amastigotes were maintained at 31°C in F-29 medium containing 20% HFBS (GIBCO BRL, Grand Island, New York) as previously reported (34). Promastigotes of this parasite line were grown at 23°C in complete Schneider's Drosophila melanogaster medium supplemented with 20% HIFBS and 10 μg/ml gentamicin. Some experiments have been repeated with L. mexicana pifanoi (MHOM/VE/57/LL1) obtained from the American Type Culture Collection (ATCC). Promastigotes of this line were cultured in complete Schneider's Drosophila medium. The amastigote forms were transformed and cultured in RPMI 1640-morpholineethanesulfonic acid (MES) (pH 5.5) medium supplemented with 20% fetal calf serum following protocols described by Debrabant et al. (14). Successful transformation of promastigotes to amastigotes was monitored by testing for the loss of reactivity to the gp46/M2 monoclonal antibody and expression of the 34-kDa form of P8 (34) in Western blots. Although data are not shown, Leishmania amazonensis (MHOM/BR/77/LTB0016) promastigotes were maintained in complete medium (Schneider's Drosophila medium; GIBCO BRL, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (GIBCO BRL, Grand Island, NY) and 10 μg/ml gentamicin at 23°C. The infectivity of parasites was maintained by periodic passage through BALB/c mice as reported previously (34). All parasites were used in the late stationary phase.
Macrophage. Immortalized RAW 264.7 murine macrophages were obtained from the ATCC. They were maintained at 37°C under a 5.5% CO2 atmosphere in RPMI 1640 complete medium supplemented with 10% heat-inactivated FBS, 100 μg/ml penicillin, 100 μg/ml streptomycin, and 10 μg/ml gentamicin.
WST-1 superoxide assay. Superoxide produced in culture was measured using a colorimetric microplate assay involving the superoxide-mediated reduction of the cell-permeable tetrazolium salt WST-1 to a water-soluble formazan precipitate (37). For the WST-1 superoxide assays, 96-well plates were seeded at 1.5 x 105 cells/well. The cultures were then incubated overnight (12 to 18 h) at 37°C at 5.5% CO2. Twenty minutes prior to infection, all macrophages were pretreated with 200 nM phorbol 12-myristate 13-acetate (PMA). WST-1 superoxide microplate assays were carried out in Hanks balanced salt solution and supplemented with 0.1% bovine serum albumin (BSA). The SOD inhibitible reaction was determined in parallel microplate wells containing 1.5 x 105 macrophages in a final volume of 0.1 ml Hanks balanced salt solution supplemented with 0.1% BSA. WST-1 (400 μM) with or without 100 U/ml SOD was added in conjunction with parasites at the time of infection. Samples were incubated at 34°C and 5% CO2 until they were taken out at the indicated intervals to be read at 438 nm using a Bio-Tek Power Wave 200 microplate reader. Superoxide production is expressed as the change in absorbance after subtracting the absorbance in SOD wells. In experiments where Sn(IV) mesoporphyrin (Sn MP) was used, the experiments were set up as above but the appropriate volume of Sn MP was added to each well to yield the desired concentrations. An identical plate of cells was prepared, to which Sn MP was added without WST-1. The absorbance values from this plate were subtracted from those obtained in the presence of WST-1. In experiments in which cobalt protoporphyrin IX (CoPP) was used, macrophages were preincubated for 2 h with a 30 μM concentration of this chemical. The macrophages were then incubated with WST-1 as discussed above. Here, too, an identical plate was set up with CoPP, and the experiment was performed without the addition of parasites. The absorbance values obtained here were subtracted from values obtained in the presence of parasites.
NBT assay. Macrophages were plated at 1 x 106 cells/dish in 60- by 15-mm petri dishes containing 12-mm round coverslips. Parasites were suspended in Dulbecco's minimal essential medium supplemented with FBS and containing 1 mg/ml NBT. The parasites were added to macrophages at a ratio of 5:1 parasites/cell. A discrete time of infection was achieved by incubating parasites with cells for 15 min at 34°C, 5% CO2 before washing the plates to remove unattached parasites and then reincubating the infected plates for the indicated time intervals. At the appropriate times, infection was terminated by the addition of 100% methanol to the culture for 5 min, followed by Giemsa (Sigma) staining of the cells on coverslips. Coverslips were mounted on glass slides with Gel Mount (Biomedia Corp., Foster City, CA). Cells were observed under differential interference contrast microscopy, and parasitized vacuoles were scored for the presence of the blue-black formazan precipitate, which forms upon the reduction of NBT by superoxide anion.
PV isolation. An enriched parasitophorous vacuole (PV) (ePV) fraction was obtained by following a previously described protocol (23). Briefly, at least 10 confluent dishes of RAW 264.7 macrophages in 100-mm petri dishes were incubated with parasites at a parasite/macrophage ratio of 4:1 and placed at 34°C and 5% CO2. After 30 min, free parasites were washed off with cold phosphate-buffered saline (PBS) and plates were returned to the incubator until the desired time. After 2 h of infection, plates were washed with cold PBS and cells were scraped into lysis buffer (20 mM HEPES, 0.5 mM EGTA, 0.25 M sucrose, and 0.1% gelatin) containing protease inhibitors (protease inhibitor cocktail; Roche, Mannheim, Germany). The cell suspension (1 ml per 10 plates) was lysed, and postnuclear supernatant (PNS) was recovered after low-speed centrifugation. The PNS was then loaded onto a sucrose step gradient (4 ml/step; 20%, 40%, 60% sucrose in gradient buffer [30 mM HEPES, 100 mM NaCl, 0.5 mM CaCl2, 0.5 mM MgCl2, pH 7.0]) and centrifuged at 700 x g for 25 min at 4°C. An ePV fraction was recovered from the 40/60% sucrose interface. The ePV fraction was then centrifuged at 12,000 x g. The protein and -hexosaminidase contents of this fraction were determined by the Bradford method.
The ePV fraction is composed of late endocytic vesicles and PVs. For affinity purification of PVs, the ePV fraction was incubated with polyclonal goat antiserum to the C-terminal tail of calnexin (Santa Cruz Biotechnology). This mixture was allowed to incubate for at least 1 h before the addition of protein G or donkey anti-goat antibody, conjugated with magnetic particles (Spherotech, Libertyville, Illinois). Antibody-bound vacuoles were positively selected with a magnet to yield the purified PV (pPV) fraction, and the supernatant fluid (V fraction) was saved for analysis. There is an enrichment of parasite molecules in the pPV; however, parasite molecules, most likely from broken parasites, are found in the remnant V fraction as well.
Western blot analysis. Equivalent protein aliquots (50 μg) of the PNS, ePV, pPV, and V fractions were separated on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted to nitrocellulose membranes according to established protocols. Magnetic particles in the pPV were removed by brief centrifugation after samples were boiled, before being loaded onto gels. For evaluation of HO-1 levels, 35 μg of lysates from infected cells, obtained at the times indicated, was resolved on 15% SDS-PAGE and then blotted onto nitrocellulose membranes. Membranes were blocked in Tris-buffered saline supplemented with 5% milk in Tris-buffered saline before incubation overnight in primary antibodies. After removal of primary antibodies and washing, membranes were incubated in the appropriate secondary antibodies conjugated to horseradish peroxidase. Washed blots were incubated with chemiluminescence (ECL kit; Amersham) reagents. Antibody reactivity was visualized by exposure of blots to X-ray film. Some blots were stripped by incubation at 50°C for 30 min in 20 mM 2-mercaptoethanol, 2% SDS, and 62.5 mM Tris-HCl, pH 6.8. They were then reprobed with other primary antibodies and processed as described above.
Immunofluorescence assays. Parasites were resuspended in complete macrophage medium and added to macrophages previously seeded on glass coverslips, at a 10:1 parasite-to-macrophage ratio. Infections were allowed to proceed at 34°C under a 5% CO2 atmosphere before washing plates to remove unattached parasites. Infections were terminated by adding an equal volume of 4% paraformaldehyde in PBS to the cultures. Coverslips were first incubated in 50 mM NH4Cl at room temperature for 15 min. Then coverslips were washed and incubated for 20 min in binding buffer (2% [wt/vol] BSA in 1x PBS) before incubation with primary antibodies. Coverslips were washed and then incubated with Alexa Fluor secondary antibodies reactive with mouse or rat immunoglobulin G (Molecular Probes, Eugene, OR). The nucleic acid dye 4',6'-diamidino-2-phenylindole dihydrochloride (DAPI; Molecular Probes) was added to the secondary antibody. Coverslips were washed and mounted on glass slides with Gel Mount (Biomedia Corp., Foster City, CA). Cells on coverslips that received secondary antibodies alone were always included in the analysis. Labeled cells were visualized and images captured on a laser scanning microscopy Pascal confocal microscope (Carl Zeiss, Inc.).
Statistics. SigmaPlot software was used to plot and analyze the data (Point Richmond, CA). Data are expressed as means ± standard errors (SE). t tests were used to analyze differences between groups. We defined statistical significance as a P value of <0.05.
RESULTS
Superoxide production by macrophages in response to L. pifanoi amastigotes. We elected to use Leishmania mexicana pifanoi, a member of the L. mexicana complex, in these studies because of available protocols to maintain axenic cultures of both the promastigote and amastigote forms of the parasite (14, 34). The capacity of these parasite forms to elicit superoxide production by macrophages was ascertained in WST-1 assays (37). Figure 1 shows that when RAW 264.7 macrophages are incubated with amastigotes, the macrophages produce negligible to low levels of superoxide. Incubation of macrophages with parasites at a ratio of 1:20 resulted in infection of greater than 50% of the macrophages. Other ratios of amastigotes to macrophages (5:1, 10:1, and 40:1) as well as antibody- and serum-opsonized amastigotes yielded similar results to those shown in Fig. 1. In contrast, significantly higher levels of superoxide were produced in parallel cultures of macrophages exposed to late-stationary-stage L. mexicana pifanoi promastigotes at 10:1 (P = 0.0004) and 20:1 parasite/macrophage ratios(P = 0.01). The response of macrophages to zymosan ingestion was also determined. Both nonopsonized zymosan (not shown) and antibody-opsonized zymosan elicited the release of superoxide as shown in Fig. 1. These results show that, whereas infection with the amastigote form of L. mexicana pifanoi barely induces the release of superoxide by macrophages, macrophages infected with the promastigote form produce significant levels of superoxide. Similar experiments have been performed with promastigotes of L. amazonensis, and we have found that they too elicit comparable responses to promastigotes of L. mexicana pifanoi. Also, infection of thioglycolate-elicited peritoneal cells from BALB/c mice yielded similar results (not shown).
The NADPH oxidase enzyme complex doesn't assemble on PVs that harbor amastigotes within infected cells. As an additional assessment of superoxide production, we determined whether the NADPH oxidase enzyme complex assembles around PVs that harbor Leishmania parasites. For these studies, we performed a widely accepted functional assay for the presence of the NADPH oxidase enzyme complex on a target membrane by assessing formazan precipitation as a result of the reduction of NBT by oxygen radicals produced by the enzyme. Representative images of NBT-positive parasitized vacuoles that harbor promastigotes and NBT-negative vacuoles that harbor amastigotes are shown in Fig. 2A and B, respectively. At 30 min postinfection, over 25% of PVs that harbor promastigotes were found to be NBT positive (Fig. 2C). There was a gradual decrease in the number of NBT-positive PVs through to the 6-h sampling point. In parallel experiments, less than 2% of vacuoles that harbored amastigotes could be scored as NBT positive. Pretreatment of macrophages with gamma interferon didn't augment the number of NBT-positive, amastigote-containing PVs (not shown).
Although most amastigote PVs were found to be NBT negative, those experiments didn't indicate at which point the assembly of the NADPH oxidase complex onto PVs is inhibited. Since formation of the NADPH oxidase complex can be initiated first by the recruitment of the membrane-bound subunits (15), we sought to determine whether the membrane subunits of the NADPH oxidase complex, particularly gp91phox, localize to amastigote PVs. Immunofluorescence staining of infected cells was performed to evaluate the distribution of gp91phox. We took advantage of the observation that after 1 h of uptake of zymosan particles or Leishmania parasites (promastigotes or amastigotes), over 95% reside in vacuoles that are Lamp1 positive. Lamp1 immunolabeling of vacuoles can therefore be used to delineate the contours of PVs. Figure 3 shows representative confocal images of macrophages infected with amastigotes (Fig. 3A, B, C, and D) or macrophages that engulfed zymosan particles (Fig. 3E, F, G, and H). Lamp1 and gp91phox staining are shown, and in the merged images, colocalization of gp91phox and Lamp1 is shown. As Fig. 3H shows, gp91phox is abundantly recruited to zymosan-containing vacuoles. In macrophages infected with amastigotes (Fig. 3D), there is a lot less gp91phox recruited to the membrane of amastigote-containing vacuoles than to zymosan-containing vacuoles. There are, however, patches of gp91phox on the PVs that harbor amastigotes. The pattern of gp91phox recruitment on promastigote-containing vacuoles is intermediate between the observations with zymosan and amastigotes (not shown). Taken together, these results confirm that the NADPH enzyme complex is minimally recruited to vacuoles that harbor Leishmania amastigotes, even though these vacuoles are sometimes reactive with antibodies to gp91phox
Full-length gp91 is absent from amastigote-infected parasitophorous vacuole fractions. Microscopic studies provide a powerful visual assessment of the distribution of molecules in cells, but they are less suitable for providing information about the structure of the molecule that is being monitored. We therefore used an additional approach to assess the recruitment of gp91phox to PVs that harbor Leishmania amastigotes. For these studies, we took advantage of an established protocol for isolating enriched PV fractions (8) and a more recently described protocol for obtaining affinity-purified PVs (23). Macrophages that had either ingested zymosan particles or been infected with amastigotes for 2 h were lysed, and subcellular fractions were obtained. Equivalent protein amounts of the PNS, ePV, pPV, and remaining V fractions were analyzed. The presence of gp91phox in these preparations was probed in Western blots using the 54.1 monoclonal antibody (5). The representative results in Fig. 4 show that the 91-kDa molecule which is reactive to monoclonal antibody 54.1, is present in the PNS, ePV, and pPV fractions obtained from macrophages that ingested zymosan particles. In contrast, gp91phox is not found in the ePV and pPV fractions obtained from amastigote-infected macrophages, even though gp91phox is detected in the PNS obtained after amastigote infections. Clearly evident though, in these fractions, is the 65-kDa endoplasmic form of gp91phox. This 65-kDa molecule was also detected with the NL7 monoclonal antibody (not shown), which is also specific for gp91phox (6).
We next probed for the presence of p22phox, the other membrane-bound subunit of the NADPH oxidase enzyme complex. Interactions between p22phox and gp91phox are essential for the stability of p22phox on intracellular membranes (15, 31). Figure 4 shows that p22phox is present in the subcellular fractions obtained from macrophages that had ingested zymosan particles. Although the PNS from amastigote-infected cells contains p22phox, much less is found in the ePV fraction and the rest of the vacuolar fractions obtained from these infected cells. This is most likely because of the absence of full-length gp91phox. We had previously found that Lamp1 is distributed throughout these subcellular fractions and can therefore be used as an internal control for protein loading (23). Taken together, these results show that, in amastigote-infected cells, the maturation of the p65 form of gp91phox to the full-length molecule is blocked. Reactivity of the 65-kDa form with antibodies to gp91phox might account for gp91phox-reactive PVs observed in confocal microscopy studies (Fig. 3).
Leishmania infection triggers an increase in HO-1 levels. The membrane-bound subunit of the NADPH oxidase enzyme, cytochrome b558, is composed of p22phox and gp91phox bound to heme. A couple of reports have shown that modulation of heme levels by either blocking their synthesis or inducing their degradation results in a decrease in gp91phox levels without affecting the abundance of p65 (15, 36). Changes in HO-1, the rate-limiting enzyme in heme degradation, was found to be a good indicator of heme degradation (36). Since gp91phox was not detected in subcellular fractions obtained from macrophages infected with amastigotes but the levels of p65 appeared to be unaffected, we performed experiments to determine whether infection of macrophages with these parasites induces heme degradation. On immunoblots, we assessed changes in the level of HO-1 after infection with Leishmania parasites or ingestion of zymosan particles. Figure 5A shows that infection with Leishmania amastigotes results in a significant increase in HO-1 levels as early as 1 h after infection and is sustained several hours postinfection. A plot of the densitometric scan from three identical experiments (Fig. 5B) confirms that the levels of HO-1 through the first 9 h postinfection are always significantly higher than the levels in resting cells (RawR). Curiously, infection with the promastigote form also results in an increase in HO-1 levels. It is noteworthy, though, that the level of HO-1 in cells exposed to amastigotes was always higher at 1 h after infection that in promastigote-infected cells. An increase in HO-1 levels in response to both parasite forms might explain why macrophages infected with promastigotes are progressively refractory to the production of superoxide (40). Ingestion of Zymosan particles resulted in minimal change in HO-1 levels after 1 h or 9 h (not shown). The level of actin in these samples was used as an internal control for sample loading (Fig. 5B).
Agents that modulate heme levels or heme activity alter superoxide production by macrophages infected with amastigotes or promastigotes. In light of the observation that infection induces HO-1, two experimental designs were employed to determine whether HO-1 plays a role in superoxide production by macrophages infected with Leishmania parasites. There are several synthetic heme analogs that can regulate heme availability either by inducing its degradation through the induction of HO-1 (CoPP) or by blocking its degradation by competitive inhibition of HO-1 (22, 24, 25). In the first design, infections were performed in the presence of tin mesoporphyrin (Sn MP), which is an inhibitor of HO-1 activity, to determine whether blocking of HO-1 activity would relieve macrophages to produce superoxide when they are exposed to amastigotes. In the second approach, macrophages were pretreated with CoPP, which induces HO-1 production. These experiments were designed to assess what the effect of higher levels of HO-1 at the time of infection would be on superoxide production by macrophages in response to Leishmania infection. Figures 6 and 7 show the results of WST-1 assays with these compounds, which were compiled from three experiments. Macrophages infected with amastigotes in the presence of 50 μM Sn MP produced significantly higher levels of superoxide than macrophages that were incubated with amastigotes alone (P = 0.0095) (Fig. 6A). Infections of macrophages with promastigotes in the presence of Sn MP did not have a significant effect (P = 0.321) on superoxide production by these macrophages (Fig. 6B).
Complementary results were obtained when macrophages were preincubated with CoPP. Figure 7 shows that superoxide production by macrophages that were preincubated for 2 h with CoPP and then incubated with promastigotes produced significantly less superoxide (P < 0.05) than macrophages that were incubated with promastigotes without preincubation with CoPP. Preincubation with CoPP also affected superoxide production by amastigote-infected macrophages: it further reduced the minimal superoxide levels that were produced in response to amastigotes. These experiments show a significant dependence of the capacity to produce superoxide on HO-1 levels. We should note that a new batch of WST-1 reagent was used in these latter experiments, which resulted in higher absorbance readings. Nonetheless, the finding that amastigote-infected macrophages produced significantly less superoxide than promastigote-infected macrophages was consistent.
Amastigotes suppress superoxide production by macrophages in coculture experiments. We wondered what effect amastigotes would have on particulate stimuli that can trigger macrophage superoxide production. To address this question, coincubation studies with amastigotes and promastigotes or amastigotes and zymosan were performed in the presence of PMA. We presumed that zymosan particles and promastigotes engage different macrophage receptors to activate superoxide production. Macrophages were incubated with promastigotes, and then in parallel cultures, increasing numbers of live or fixed amastigotes were added to the promastigote infection. WST-1 assays that assessed superoxide production showed that when the increasing number of amastigotes was used, there was a reduction of superoxide production by macrophages in response to promastigote infection (Fig. 8a). In the presence of a 30:1 amastigote-to-macrophage ratio, promastigote-infected macrophages released 40% less superoxide (P < 0.03) than in infections with promastigotes alone. To rule out the likelihood that amastigotes were merely outcompeting promastigotes for interactions with macrophages, heat-killed (not shown) or paraformaldehyde-fixed amastigotes were added to promastigote cultures in comparable numbers to the live parasite experiments. Figure 8b shows that paraformaldehyde-fixed parasites had no significant effect on promastigote-elicited superoxide production.
We also determined the effect of amastigotes on zymosan elicitation of superoxide production by macrophages. As Fig. 8C shows, coaddition of amastigotes at a ratio of 20:1 (parasite to macrophages) with zymosan (10:1) resulted in greater that 80% reduction in superoxide production by the macrophages. For comparison, promastigotes and zymosan particles were added together on macrophages, and superoxide production was assessed. Unlike the scenario with amastigotes, coaddition of zymosan particles with promastigotes on macrophages resulted in greater levels of superoxide production. These results show that Leishmania amastigotes can inhibit superoxide production elicited by different stimuli.
DISCUSSION
Reactive oxygen radicals are some of the potent antimicrobial responses that phagocytes elaborate to limit the growth of microorganisms. To survive and thrive, intracellular pathogens have to either avoid elicitation of oxygen radicals or resist their effects. In this report, we provide evidence that supports the finding that murine macrophages produce minimal levels of superoxide when they internalize the amastigote form of parasites of the L. mexicana complex. Three experimental approaches were employed to elucidate the underlying mechanism(s) that enables this phenomenon. First, in a histochemical assay using nitroblue tetrazolium, which is reduced to a formazan precipitate on membranes that contain the functional NADPH oxidase enzyme, we found that very few vacuoles that harbor the amastigote form of the parasite contain the NADPH oxidase complex (Fig. 2). This was contrasted with infections with the promastigote form, where 23% of those vacuoles were positive for the NADPH oxidase complex. The advantage of this assay is that it is minimally affected by the superoxide scavenger, super oxide dismutase, which is located in the mitochondria and the cytosol (12). Second, immunolocalization studies of gp91phox revealed that even though the NADPH oxidase enzyme complex isn't present on amastigote-containing vacuoles, some of these vacuoles are positive for gp91phox (Fig. 3). This raised the possibility that gp91phox molecules that are recruited to these vacuoles might not always be functional. Third, whereas subcellular fractions enriched with vacuoles that harbor zymosan particles contain detectable levels of gp91phox, comparable subcellular fractions from Leishmania amastigote-infected cells were devoid of the mature form of gp91phox, even though detectable levels of gp91phox were found in the postnuclear fraction. Instead, the precursor form of gp91phox, p65, is readily detected in these subcellular fractions. Taken together, these experiments suggested that amastigotes avoid the elicitation of superoxide by promoting activities that result in the blockage of the maturation of gp91phox and thus prevent the formation of functional enzyme complexes in infected cells.
Several microorganisms have been shown to inhibit the production of superoxide by macrophages and neutrophils. Some examples are Yersinia enterocolitica (41), Coxiella burnetii (1), Chlamydia trachomatis (38), and the agent of human granulocytic ehrlichiosis (HGE agent) (7, 28), just to name a few. These organisms suppress the release of superoxide when neutrophils are stimulated by known inducers of superoxide. From this group of organisms, the HGE agent was found to be somewhat unique because it suppresses superoxide release when neutrophils are exposed to stimuli that induce superoxide production through diverse mechanisms (29, 30). The capacity to inhibit superoxide release induced by multiple stimuli is an indication that the effect of this organism is at a point where several signaling pathways converge. In this regard, our observations with Leishmania amastigotes are most similar to the observations with the HGE agent. In Fig. 8, we showed that Leishmania amastigotes, too, suppress superoxide production by macrophages that were exposed to zymosan particles and promastigotes in the presence of PMA.
The regulation of superoxide production by microorganisms can occur at several points in their interaction. In this study, we focused on activities that affect the posttranslational processing of gp91phox, which eventually affects the assembly of the enzyme subunits. However, some studies, particularly those employing the HGE agent, have investigated the likelihood that suppression of superoxide production is a consequence of a targeted block in the transcription of gp91phox (2, 39). Those observations were made after a 5-day incubation of the HGE agent with neutrophils (2). Other studies in that infection system that have investigated the suppression of superoxide production in a much shorter frame (within 2 h) have found no changes in the level of gp91phox (29), even though the inhibition of superoxide production is significant by then. This suggests that modulating the transcription of gp91phox might not be a relevant strategy for the immediate survival of internalized organisms.
Gp91phox is synthesized as a 56-kDa molecule in the endoplasmic reticulum, where it undergoes limited glycosylation and transforms into a 65-kDa protein. Further processing of p65 occurs beyond the endoplasmic reticulum, where it is additionally glycosylated and acquires two heme molecules (31). Only the full-length gp91phox can stably associate with p22, the other membrane-bound subunit of the NADPH oxidase molecule (31). It was recently shown that induction of heme degradation with succinyl acetone, results in inhibition of gp91phox maturation into the fully glycosylated form, without affecting the levels of p65 (15). The addition of heme reversed the effect of succinyl acetone (43). In our studies, Western blot analysis of gp91phox distribution in subcellular fractions of amastigote-infected cells (Fig. 4) indicated that these parasites cause a comparable blockage of gp91phox maturation, as does succinyl acetone. This led us to hypothesize that infection with amastigotes results in an increase in HO-1, as does succinyl acetone, which results in impaired gp91phox maturation and limited superoxide production.
HO-1, also called HSP 32, is the rate-limiting enzyme in the degradation of heme (25). It catalyzes the conversion of heme to biliverdin. HO-1 is induced by a wide variety of stimuli. HO-1 levels can be rapidly induced, and maximal levels can be reached in 1 h (17). It is obvious from the studies presented here that both the promastigote and amastigote forms of the parasite induce macrophages to upregulate their levels of HO-1. However, it appears that amastigotes trigger a much quicker rise in HO-1 than is observed with promastigotes (Fig. 5). There is some evidence that promastigote and amastigote forms of Leishmania exhibit differences in heme metabolism (synthesis and degradation) (9, 35), but it is unlikely that such activities within the parasite contribute to the rapid effect on macrophage HO-1 levels.
Although we cannot rule out the likelihood that amastigotes might suppress superoxide production through alternate mechanisms, we are confident that the induction of HO-1 plays a significant role in the suppression of superoxide production. Two approaches were employed to fully implicate the induction of HO-1 in the underlying mechanism employed by amastigotes to avoid superoxide production. In the first, when HO-1 activity was blocked with competitive inhibitors, more superoxide is produced by macrophages in response to amastigote infection. In the second approach, induction of HO-1 prior to infection resulted in reduced superoxide production in response to infection. Metalloporphyrins, such as zinc protoporphyrin and Sn MP, are competitive inhibitors of HO-1 activity (22, 24) that have been used in the treatment of infants with neonatal jaundice, where the accumulation of bilirubin, a downstream metabolite of heme, is believed to be part of the pathophysiology of disease (20). We observed that Sn MP significantly reversed the inhibitory effect of amastigote infection on superoxide production (Fig. 6). Interestingly, Sn-MP didn't have a significant effect on superoxide production in promastigote infections. But preincubation with CoPP, which induces HO-1 production, resulted in a reduction of superoxide production in response to infection with promastigotes. This later result suggests that there might be an HO-1 threshold at which point its effect on heme levels has a significant effect on heme-containing compounds like gp91phox. So even though there is only a small difference in the HO-1 levels induced by promastigotes and amastigotes after 1 h (Fig. 5), that difference might be sufficient to result in a greater suppression of superoxide production by amastigotes.
The finding that Leishmania infections in the presence of metalloporphyrins result in increased superoxide production is an important observation. Since it has previously been shown that Leishmania parasites are susceptible to superoxide-mediated killing (19, 33), a strategy that augments superoxide in the vicinity of these parasites might result in the control of Leishmania infections. There is already extensive experience with using metalloporphyrins in humans, so it should be feasible to adapt them for use in this disease. Initial studies to assess their usefulness in controlling Leishmania infections are being considered in experimental models of the disease.
Amastigotes block superoxide production by other stimuli. Coincubation of amastigotes with promastigotes or zymosan particles muted the response of macrophages to these potent stimuli of superoxide production (Fig. 7). Fixed or heat-killed amastigotes had minimal effects on the macrophage response to promastigotes. This would suggest that live amastigotes actively alter processes within the macrophage that results in their refractoriness to otherwise potent stimuli. Since it is most probable that promastigotes, amastigotes, and zymosan particles engage different signaling pathways when they are engulfed by cells, the suppression of superoxide production of these stimuli suggest that the effect of amastigotes occurs at a point where all the signals converge.
Taken together, the studies presented here show that amastigotes not only avoid the elicitation of superoxide production during their internalization but also render the macrophages refractory to superoxide production by otherwise potent stimuli. This effect on heme metabolism is an evasion strategy that might be employed by other organisms.
Additionally, while the manuscript was being revised, a paper by Chauveau et al. (11) showed that induction of HO-1 by CoPP pretreatment resulted in a reduction in superoxide production by dendritic cells in response to lipopolysaccharide stimulation.
ACKNOWLEDGMENTS
We thank Algirdas J. Jesaitis and James Burritt at Montana State University for generous gifts of anti-gp91phox and p22 monoclonal antibodies (54.1, NL7, and 44.1). We also thank Mary Wilson for reading the manuscript and for suggestions on improving it.
This study was funded by grant AI048739 from the National Institutes of Health.
REFERENCES
1. Baca, O. G., M. J. Roman, R. H. Glew, R. F. Christner, J. E. Buhler, and A. S. Aragon. 1993. Acid phosphatase activity in Coxiella burnetii: a possible virulence factor. Infect. Immun. 61:4232-4239.
2. Banerjee, R., J. Anguita, D. Roos, and E. Fikrig. 2000. Cutting edge: infection by the agent of human granulocytic ehrlichiosis prevents the respiratory burst by down-regulating gp91phox. J. Immunol. 164:3946-3949.
3. Blos, M., U. Schleicher, F. J. Soares Rocha, U. Meissner, M. Rollinghoff, and C. Bogdan. 2003. Organ-specific and stage-dependent control of Leishmania major infection by inducible nitric oxide synthase and phagocyte NADPH oxidase. Eur. J. Immunol. 33:1224-1234.
4. Brandonisio, O., M. A. Panaro, R. Marzio, A. Marangi, S. M. Faliero, and E. Jirillo. 1994. Impairment of the human phagocyte oxidative responses caused by Leishmania lipophosphoglycan (LPG): in vitro studies. FEMS Immunol. Med. Microbiol. 8:57-62.
5. Burritt, J. B., M. T. Quinn, M. A. Jutila, C. W. Bond, and A. J. Jesaitis. 1995. Topological mapping of neutrophil cytochrome b epitopes with phage-display libraries. J. Biol. Chem. 270:16974-16980.
6. Burritt, J. B., T. R. Foubert, D. Baniulis, C. I. Lord, R. M. Taylor, J. S. Mills, T. D. Baughan, D. Roos, C. A. Parkos, and A. J. Jesaitis. 2003. Functional epitope on human neutrophil flavocytochrome b558. J. Immunol. 170:6082-6089.
7. Carlyon, J. A., D. Abdel-Latif, M. Pypaert, P. Lacy, and E. Fikrig. 2004. Anaplasma phagocytophilum utilizes multiple host evasion mechanisms to thwart NADPH oxidase-mediated killing during neutrophil infection. Infect. Immun. 72:4772-4783.
8. Chakraborty, P., S. Sturgill-Koszycki, D. G. Russell. 1994. Isolation and characterization of pathogen-containing phagosomes. Methods Cell Biol. 45:261-276.
9. Chang, C. S., and K. P. Chang. 1985. Heme requirement and acquisition by extracellular and intracellular stages of Leishmania mexicana amazonensis. Mol. Biochem. Parasitol. 16:267-276.
10. Channon, J. Y., M. B. Roberts, J. M. Blackwell. 1984. A study of the differential respiratory burst activity elicited by promastigotes and amastigotes of Leishmania donovani in murine resident peritoneal macrophages. Immunology 53:345-355.
11. Chauveau, C., S. Remy, P. J. Royer, M. Hill, S. Tanguy-Royer, F. X. Hubert, L. Tesson, R. Brion, G. Beriou, M. Gregoire, R. Josien, M. C. Cuturi, and I. Anegon. 2005. Heme oxygenase-1 expression inhibits dendritic cell maturation and proinflammatory function but conserves IL-10 expression. Blood 106:1694-1702. (First published 26 May 2005; doi:10.1182/blood-2005-02-0494.)
12. Crapo, J. D., T. Oury, C. Rabouille, J. W. Slot, and L.-Y. Chang. 1992. Copper, zinc superoxide dismutase is primarily a cytosolic protein in human cells. Proc. Natl. Acad. Sci. USA 89:10405-10409.
13. Da Silva, R. P., B. F. Hall, K. A. Joiner, and D. L. Sacks. 1989. CR1, the C3b receptor, mediates binding of infective Leishmania major metacyclic promastigotes to human macrophages. J. Immunol. 143:617-622.
14. Debrabant, A., M. B. Joshi, P. F. Pimenta, and D. M. Dwyer. 2004. Generation of Leishmania donovani axenic amastigotes: their growth and biological characteristics. Int. J. Parasitol. 34:205-217.
15. DeLeo, F. R., J. B. Burritt, L. Yu, A. J. Jesaitis, M. C. Dinauer, and W. M. Nauseef. 2000. Processing and maturation of flavocytochrome b558 include incorporation of heme as a prerequisite for heterodimer assembly. J. Biol. Chem. 275:13986-13993.
16. Descoteaux, A., and S. J. Turco. 1993. The lipophosphoglycan of Leishmania and macrophage protein kinase C. Parasitol. Today 9:468-471.
17. Ewing, J. F., S. N. Haber, and M. D. Maines. 1992. Normal and heat-induced patterns of expression of heme oxygenase-1 (HSP32) in rat brain: hyperthermia causes rapid induction of mRNA and protein. J. Neurochem. 58:4023-4029.
18. Gantt, K. R., T. L. Goldman, M. A. Miller, M. L. McCormick, S. M. B. Jeronimo, E. T. Nascimento, B. E. Britigan, and M. E. Wilson. 2001. Oxidative responses of human and murine macrophages during phagocytosis of Leishmania chagasi. J. Immunol. 167:893-901.
19. Haidaris, C. G., and P. F. Bonventre. 1982. A role for oxygen-dependent mechanisms in killing of Leishmania donovani tissue forms by activated macrophages. J. Immunol. 129:850-855.
20. Ip, S., M. Chung, J. Kulig, R. O'Brien, R. Sege, S. Glicken, M. J. Maisels, and J. Lau. 2004 An evidence-based review of important issues concerning neonatal hyperbilirubinemia. Pediatrics 114:e130-e153.
21. Kantengwa, S., I. Muller, J. Louis, and B. S. Polla. 1995. Infection of human and murine macrophages with Leishmania major is associated with early parasite heat shock protein synthesis but fails to induce a host cell stress response. Immunol. Cell Biol. 73:73-80.
22. Kappas, A., and G. S. Drummond. 1985. Metalloporphyrins: a class of compounds of pharmacological interests. BioEssays 3:256-259.
23. Kima, P. E., and W. Dunn. 2005. Exploiting calnexin expression on phagosomes to isolate Leishmania parasitophorous vacuoles. Microb. Pathog. 38:139-145.
24. Maines, M. D. 1981. Zinc-protoporphyrin is a selective inhibitor of heme oxygenase activity in the neonatal rat. Biochim. Biophys. Acta 673:339-350.
25. Maines, M. D. 1997. The heme oxygenase system: a regulator of second messenger gases. Annu. Rev. Pharmacol. Toxicol. 37:517-554.
26. Mallinson, D. J., and G. H. Coombs. 1989. Interaction of Leishmania metacyclics with macrophages. Int. J. Parasitol. 19:647-656.
27. McMahon-Pratt, D., and J. Alexander. 2004. Does the Leishmania major paradigm of pathogenesis and protection hold for New World cutaneous leishmaniasis or the visceral disease Immunol. Rev. 201:206-224.
28. Mott, J., and Y. Rikihisa. 2000. Human granulocytic ehrlichiosis agent inhibits superoxide anion generation by human neutrophils. Infect. Immun. 68:6697-6703.
29. Mott, J., Y. Rikihisa, and S. Tsunawaki. 2002. Effects of Anaplasma phagocytophila on NADPH oxidase components in human neutrophils and HL-60 cells. Infect. Immun. 70:1359-1366.
30. Murray, H. W., and C. F. Nathan. 1999. Macrophage microbicidal mechanisms in vivo: reactive nitrogen versus oxygen intermediates in the killing of intracellular visceral Leishmania donovani. J. Exp. Med. 189:741-746.
31. Nauseef, W. M. 2004. Assembly of the phagocyte NADPH oxidase. Histochem. Cell Biol. 122:277-291.
32. Passwell, J. H., R. Shor, J. Smolen, and C. L. Jaffe. 1994. Infection of human monocytes by Leishmania results in a defective oxidative burst. Int. J. Exp. Pathol. 75:277-284.
33. Reiner, N. E., and J. W. Kazura. 1982. Oxidant-mediated damage of Leishmania donovani promastigotes. Infect. Immun. 36:1023-1027.
34. Soong, L., S. M. Duboise, P. Kima, and D. McMahon-Pratt. 1995. Leishmania pifanoi amastigote antigens protect mice against cutaneous leishmaniasis. Infect. Immun. 63:3559-3566.
35. Srivastava, P., G. D. Sharma, K. K. Kamboj, A. K. Rastogi, and W. C. Pandey. 1997. Heme metabolism in promastigotes of Leishmania donovani. Mol. Cell. Biochem. 171:65-68.
36. Taille, C., J. El-Benna, S. Lanone, M. C. Dang, E. Ogier-Denis, M. Aubier, and J. Boczkowski. 2004. Induction of heme oxygenase-1 inhibits NAD(P)H oxidase activity by down-regulating cytochrome b558 expression via the reduction of heme availability. J. Biol. Chem. 279:28681-28688.
37. Tan, A. S., and M. V. Berridge. 2000. Superoxide produced by activated neutrophils efficiently reduces the tetrazolium salt, WST-1 to produce a soluble formazan: a simple colorimetric assay for measuring respiratory burst activation and for screening anti-inflammatory agents. J. Immunol. Methods 238:59-68.
38. Tauber, A. I., N. Pavlotsky, J. S. Lin, and P. A. Rice. 1989. Inhibition of human neutrophil NADPH oxidase by Chlamydia serovars E, K, and L2. Infect. Immun. 57:1108-1112.
39. Thomas, V., S. Samanta, C. Wu, N. Berliner, and E. Fikrig. 2005. Anaplasma phagocytophilum modulates gp91phox gene expression through altered interferon regulatory factor 1 and PU.1 levels and binding of CCAAT displacement protein. Infect. Immun. 73:208-218.
40. Uchmuller-Rouiller, Y., and J. Mauel. 1987. Impairment of the oxidative metabolism of mouse peritoneal macrophages by intracellular Leishmania spp. Infect. Immun. 55:587-593.
41. Visser, L. G., E. Seijmonsbergen, P. H. Nibbering, P. J. van den Broek, and R. van Furth. 1999. Yops of Yersinia enterocolitica inhibit receptor-dependent superoxide anion production by human granulocytes. Infect. Immun. 67:1245-1250.
42. Vouldoukis, I., P.-A. Becherel, V. Riveros-Moreno, M. Arock, O. da Silva, P. Debre, D. Mazier, and M. D. Mossalayi. 1997. Interleukin-10 and interleukin-4 inhibit intracellular killing of Leishmania infantum and Leishmania major by human macrophages by decreasing nitric oxide generation. Eur. J. Immunol. 27:860-865.
43. Yu, L., F. R. DeLeo, K. J. Biberstine-Kinkade, J. Renee, W. M. Nauseef, and M. C. Dinauer. 1999. Biosynthesis of flavocytochrome b558. gp91(phox) is synthesized as a 65-kDa precursor (p65) in the endoplasmic reticulum. J. Biol. Chem. 274:4364-4369.
44. Yu, L., L. Zhen, and M. C. Dinnauer. 1997. Biosynthesis of the phagocyte NADPH oxidase cytochrome b558. J. Biol. Chem. 272:27288-27294.(Nam-Kha Pham, Jennifer Mo)