Plasmodium berghei Resists Killing by Reactive Oxygen Species
http://www.100md.com
感染与免疫杂志 2005年第10期
La Jolla Bioengineering Institute, La Jolla, California 92037
ABSTRACT
Reactive oxygen species (ROS) are widely believed to kill malarial parasites. C57BL/6 mice injected with P. berghei inocula incubated with supraphysiological doses of NO (150 μM) or with peroxynitrite (220 μM), however, exhibited parasitemia similar to that seen with those given control inocula, and there was no difference in disease development. Only treatment of inocula with NO doses nearing saturation (1.2 mM) resulted in no detectable parasitemia in the recipients; flow cytometric analysis with a vital dye (hydroethidine) indicated that 1.5 mM NO lysed the erythrocytes rather than killing the parasites. The hemoglobin level in the inocula was about 8 μM; the hemoglobin was mainly oxyhemoglobin (oxyHb) (96%), which was converted to methemoglobin (>95%) after treatment with 150 μM NO. The concentrations of 150 μM of NO and 220 μM of peroxynitrite were far in excess of the hemoglobin concentration (8 μM), and yet no parasite killing was detected. We therefore conclude that hemoglobin protects Plasmodium parasites from ROS, but the parasite likely possesses intrinsic defense mechanisms against ROS.
INTRODUCTION
Malaria, a reemerging disease (20) caused by the parasite of the genus Plasmodium, remains refractory to the development of a vaccine in part due to incomplete understanding of the mechanism(s) underlying parasite killing by the immune system. It is generally accepted that reactive oxygen species (ROS), including nitric oxide (NO), superoxide, and peroxynitrite, kill intraerythrocytic malarial parasites (5, 8, 29). The most cited mechanism for parasite killing is that acute Plasmodium infection induces gamma interferon-producing Th1 cells, which in turn activate macrophages to secrete parasiticidal NO and ROS (29, 30).
We propose, however, that the blood stage Plasmodium parasite is virtually immune to the cytotoxic effects of NO and ROS as a consequence of hemoglobin (Hb) NO scavenging and ROS suppression within red blood cells (RBCs). The Plasmodium parasite is surrounded by hemoglobin through most of its asexual blood cycle, because it resides within a parasitophorous vacuole inside erythrocytes. Plasmodium falciparum parasites rupture erythrocytes, releasing progeny merozoites, which invade new RBCs after completing their 48-hour blood stage cycle. This extracellular excursion constitutes a brief period in which the parasite is in principle vulnerable to higher ROS concentrations induced by the infection. However, the disruption of the RBC membrane inevitably releases molecular hemoglobin into the circulation, enhancing ROS scavenging; thus, both inside and outside the red cells the parasite is protected from ROS because ROS are scavenged by hemoglobin.
Malaria, therefore, is fundamentally different from most infections, because the parasite is surrounded by hemoglobin and can evade the ROS-based protective mechanism as a consequence of ROS quenching by Hb (2), an antioxidant mechanism that has been overlooked. Although Hb's heme group can undergo redox transitions to higher oxidation states and it can auto-oxidize naturally to form methemoglobin (metHb) and superoxide ions, this oxidation process is controlled within RBCs by the metHb reductase system (1). The biochemical basis for the in vivo hemoglobin redox reactions has been extensively reviewed by Alayash (2).
To test the hypothesis that the Plasmodium parasite evades the ROS-based protective mechanisms because the parasite is surrounded by hemoglobin, we analyzed NO and ROS effects on Plasmodium berghei ex vivo, with particular attention to the molecular state of hemoglobin. Although there are differences between human and experimental malaria which limit extrapolation to the human condition, ex vivo treatments of P. berghei can assess the effect of NO and ROS on common features of parasite viability and replication as well as its ability to elicit disease.
MATERIALS AND METHODS
Evaluation of NO solution bioactivity. (i) Nitric oxide stock preparation. Phosphate-buffered saline (PBS) solution (Invitrogen, Carlsbad, CA) (pH 7.4) was degassed under vacuum overnight and placed in line between a 2 M NaOH solution and a 1 M KMnO4 solution. Prepurified grade 4.8 (99.998% pure) nitrogen was bubbled through the system for 30 min followed by UHP nitric oxide (Airgas, San Diego, CA) (99.5% pure) for 5 min (13). A saturated solution of NO has a concentration of 1.8 mM at room temperature and 1 atm (13), which we verified using the amperometric inno-T nitric oxide system (Innovative Instruments, Inc.). NO stock solutions were prepared in an anaerobic chamber filled with grade 4.8 nitrogen gas, 2 μl or 5 μl (1:10,000 and 1:4,000 dilution, respectively) was injected into a stirred 20 ml sample of PBS, and the resultant NO concentrations were monitored as a function of time and were 176.9 ± 6.6 nM for the 10,000 dilution and 435.6 ± 2.3 nM for the 4,000 dilution.
(ii) Hemoglobin solution preparation and NO concentration analysis. Hemoglobin solutions were prepared by lysing freshly isolated murine RBCs. Blood was obtained from anesthetized C57BL/6 mice (Jackson Laboratories) and centrifuged at 500 x g for 10 min to pellet the erythrocytes. The supernatant was aspirated, and then the cells were lysed with 0.9 ml distilled water. The solution was later osmotically balanced with the addition of 0.1 ml 10x PBS. The oxyhemoglobin (oxyHb) and methemoglobin concentration of this stock was determined using the Winterbourn spectrophotometric method (35). The oxyhemoglobin assay for NO (13, 22) was used to verify the effective dose of NO treatments. Hemoglobin solutions of 1.2 mM, 600 μM, 60 μM, and 20 μM were prepared and were treated three times using Hamilton syringes with 0.7 ml, 125 μl, 12.5 μl, and 1.25 μl of saturated NO solutions, respectively. Following each injection, the change in oxyHb level was determined by monitoring the change in absorption at 577 nm with a μQuant spectrophotometer (BioTek, Winooski, VT).
Infection of mice with ROS-treated P. berghei. Plasmodium berghei ANKA was stored as a frozen stabilate. The stabilate was thawed and then injected into a source mouse to generate an inoculum for the experiments. Blood from the source mouse was diluted with PBS to prepare aliquots of the inoculum. In the first series of experiments, the blood was diluted 1:6 in PBS upon collection from the source mouse. The number of RBC per milliliter was determined by using a hemocytometer. This value and the percent parasitemia were used to calculate the final dilution necessary to obtain 7.5 x 106 parasitized erythrocytes (pRBCs) in 1.5 ml PBS (5 x 106 pRBC/ml), and this dilution (100-fold) was then achieved. In subsequent experiments, the blood cells were kept in a 1:6 dilution of PBS and then diluted to the final concentration just before they were exposed to either NO or SIN-1 (3-morpholino-sydnonimine) as described below. A 200-μl volume of the NO- or SIN-1-exposed inoculum (1 x 106 pRBC) was injected intraperitoneally (i.p.) into groups of three to five C57BL/6J mice for each treatment group. The mice were obtained from Jackson Laboratories (Bar Harbor, ME) at 6 to 8 weeks of age and were injected with parasitized RBCs (pRBCs) when they were 8 to 10 weeks of age. The animals were housed in microisolator cages and provided food ad libitum. Parasitemia was assessed in each group of mice by counting the number of infected erythrocytes in Giemsa-stained thin blood films. The effects of NO treatments on the survival of severe malaria were also assessed. The P. berghei model of malaria is a very reproducible model of cerebral malaria in which virtually all C57BL/6J mice develop neurologic impairment and obtundation and then succumb to infection between days 6 and 8 of infection (14, 15). The Institutional Animal Care and Use Committee committee at the La Jolla Bioengineering Institute approved all procedures.
(i) Nitric oxide treatment. Experiments were carried out in a water-jacketed reaction chamber (WPI), preheated to 37°C, which was placed inside a custom anaerobic chamber that was degassed with grade 4.8 (99.998% pure) nitrogen for 1 h. Inoculum samples containing 7.5 million pRBCs were pipetted into the reaction chamber and treated with a bolus of nitric oxide stock solution (1,250 μl, 1,000 μl, 125 μl, 12.5 μl, or 1.25 μl) delivered via a gas-tight Hamilton syringe. The final sample volume was 1.5 ml, contained 7.5 million pRBC, and had an NO concentration of 1.5 mM, 1.2 mM, 150 μM, 15 μM, or 1.5 μM. The chamber was capped to minimize NO escape from the solution into the gas headspace, and the solution was maintained at 37°C for 10 min. The control treatment consisted of pipetting 1,250 μl of degassed PBS into the 250 μl inoculum sample containing 7.5 million pRBC. At the end of the treatment period, the treated inoculum was aspirated from the sample chamber via syringe, and 200 μl of the sample (1 million pRBC per sample) was injected i.p. into mice. The oxyhemoglobin content of inoculum was determined by lysing the erythrocytes and analyzing the lysate by use of the Cripps method (9).
(ii) NO bioactivity assessment. The inoculum was treated with 1.5 mM NO, 150 μM NO, or PBS, as described above. Instead of splitting the inoculum into samples and injecting the samples into mice, it was then centrifuged and lysed, as described above, and the Winterbourn spectrophotometric assay was used to determine oxyHb, metHb, and choleglobin levels.
(iii) SIN-1 treatment. A SIN-1 stock solution (250 mM) was prepared by dissolving 50 mg SIN-1 chloride (Alexis, San Diego, CA) in PBS and immediately freezing the solution. SIN-1 treatments consisted of adding 8 μl of SIN-1 stock to 1 ml PBS and mixing this resultant 2 mM SIN-1 solution with 1 ml of inoculum containing 10 million pRBC in PBS, yielding 2 ml of 1 mM SIN-1 solution with 5 million pRBC/ml and production rates of superoxide and nitric oxide of 7.0 μM/min and 3.7 μM/min, respectively (17). The treated inoculum was maintained in room air for either 1 h at room temperature or 10 min at 37°C. Following treatment, the inoculum was centrifuged at 500 x g for 10 min and resuspended in 2 ml of PBS, and 200 μl samples (1 million pRBC) were injected i.p. into mice as described above.
Flow cytometric assessment of parasite viability and permeability of RBCs. Inoculum samples were labeled with the vital dye hydroethidine immediately after NO treatment or following a 4-h incubation in fetal bovine serum at 37°C by the use of a previously described technique (10). Hydroethidine (Molecular Probes) at 10 mg/ml in dimethyl sulfoxide (Invitrogen Corporation, Carlsbad, CA) was diluted 1:600 in PBS. Parasite inocula (10 million) treated with 1.5 mM NO or PBS were washed once in PBS, resuspended in 0.5 ml of the hydroethidine solution, and incubated for 20 min at 37°C in the dark. Following the incubation, all hydroethidine-stained samples were washed twice in PBS and resuspended in 1 ml of PBS. Propidium iodide was added to replicates of the treated inocula to assess the number of erythrocytes that were permeable. Spherotech beads were added to all samples prior to being processed using a FACSCalibur system (Beckton Dickinson, San Diego, CA) to assess cell numbers. These beads are distinguished from cells on the basis of their unique forward and side scatter and fluorescence on all three channels. The acquisition of data was performed using the CellQuest (Becton-Dickinson) program on 10,000 cells identified by characteristic forward and side scatter. The Attractors (Becton-Dickinson) program was used to analyze the data. Cell numbers were determined using the following formula:
Statistical analysis. Analysis of variance with the Statview program (SAS Institute, Cary, NC) and Fisher's post hoc test was performed to statistically compare all parasitemia with a P value cutoff of 0.05. The means and standard deviations of the results are reported in the text and figures.
RESULTS
Verification of nitric oxide stock concentration. The decay of NO concentration with time was monitored with a NO amperometric method, showing that our stock was saturated with NO and that dilutions of the stock provided reproducible treatments of NO. The recording of NO levels in the solution diluted 1:4,000 showed that the exponential decay constant for NO in the anaerobic chamber was 20.9 ± 0.6 min. It was not possible to eliminate the headspace above the solution, so the most-rapid loss of NO was due to NO diffusion from the solution into the atmosphere. This NO diffusion was minimized in the capped reaction chamber used in the treatment of parasite inocula, so we estimate that the NO decay time constant for parasite treatments is at least twice that measured (41.8 ± 1.2 min). Either time constant indicates that most of the NO was present during the 10-min treatment period in the anaerobic chamber.
The amount of NO delivered by our treatments and by the treatment variations was determined using the oxyhemoglobin NO assay because this is the NO experienced by the RBC-based parasite, and concentrations were beyond the operating range of our NO electrode. The Hb solutions were treated with the selected NO doses, and then the amount of bioactive NO actually delivered was assessed by spectroscopically monitoring the reduction in oxyHb (13, 22). Before treatment, most (90.7 ± 2.5%) of the Hb in the test solution was oxyHb. Measurement of oxyHb concentration following treatment showed that the selected doses of NO were similar to the actual concentrations for bioactive NO (Fig. 1B).
Almost-saturated solutions of nitric oxide are needed to impair the viability of P. berghei parasites. There was no detectable effect of NO on parasitemia or development of disease in groups of mice injected i.p. with 1 x 106 P. berghei pRBCs treated with either 15 μM or 1.5 μM NO compared to the control group results (Fig. 2A). In this experiment, blood was immediately diluted (1:600) to obtain the parasite inoculum (7.5 x106 pRBC/1.5 ml) and was kept diluted in PBS for about 2 h before the injection of 1 x 106 pRBCs into the mice. Mice injected with parasitized RBCs treated with 150 μM NO showed a modest but not significant decrease in parasitemia at days 4 and 6 (P = 0.17 and P = 0.63, respectively) and slightly delayed development of experimental cerebral malaria (Fig. 2B). Thus, there appeared to be an increase in survival in groups of mice injected with parasitized RBCs treated with 150 μM NO (Fig. 2B). Only mice infected with parasitized RBCs treated with 1.5 mM NO exhibited no viable parasites and consequently no development of disease. This experiment was repeated twice with similar results.
We assessed the amount of Hb in the parasitized erythrocyte sample and assessed the state of the Hb within the erythrocytes prior to our NO treatment. The oxyheme concentration of the samples assessed by the Cripps spectrophotometric method was 29.3 ± 1.4 μM. Thus, the samples treated with 150 μM and 1.5 mM NO had sufficient NO in principle to overcome the Hb quenching of NO.
We repeated the above-described experiment except that instead of diluting the blood immediately to its final concentration (7.5 x106 pRBC/1.5 ml) for NO treatment, the blood was kept at a 1:6 dilution and then diluted further (1:100) to the final inoculum concentration just before exposure to NO. Mice were injected i.p. with 1 x 106 pRBCs from the inoculum immediately after it was exposed to NO. We performed this assay because hydroethidine results suggested that prolonged incubation in PBS alone lowered parasite viability compared with the results seen with parasite samples kept in a 1:6 dilution and then diluted immediately before use; this suggests that the lack of nutrients may have affected parasite viability. Under the new condition of parasite storage prior to NO treatment, no change in parasite viability or disease-eliciting capability was now detected at the 150 μM NO concentration (Fig. 2C). This observation indicated that a weakening of the parasite by prolonged (1 to 2 h) incubation in the absence of serum had occurred, rendering it artificially more susceptible to NO killing. The 1.2 mM NO treatment resulted in no viable parasites, and no infected mice developed disease (Fig. 2D). The oxyheme concentration in the inoculum samples was 45.6 ± 1.3 μM, indicating that the 150 μM and 1.2 mM NO treatments should have overcome the NO-quenching capacity of the oxyHb. We repeated this experiment twice with similar results. In all cases, the concentrations used for the NO treatments (1.2 mM and 150 μM) were far in excess of the oxyhemoglobin concentration of the samples.
Nitric oxide is bioactive following diffusion into erythrocytes. Because no parasite killing was detected even when the NO treatment concentration (150 μM) far exceeded the sample hemoglobin concentration (30 to 45 μM), it is possible that sufficient NO did not diffuse into the RBC during the course of the 10-min treatments. In fact, Liao and coworkers (33) have proposed that diffusional barriers slow the kinetics of NO consumption by RBCs, thereby explaining why NO consumption by free Hb is about 1,000-fold greater than that seen with the same amount of Hb packaged in RBCs. To verify that on the 10-min time scale of our treatments the diffusional barriers did not markedly inhibit NO bioavailability within the erythrocyte and that the parasite was exposed to NO, we determined the Hb oxidation state before and immediately after NO treatment of the experimental samples. Before NO treatment and placement in the anaerobic chamber, the samples were mainly oxyHb (96.1 ± 1.1%), with little metHb (3.9 ± 1.1%) or choleglobin (0.0 ± 0.0%), a denatured form of Hb. Following NO treatment of the parasitized erythrocytes, most of the oxyHb (95%) was converted into either metHb or choleglobin (Fig. 3). Maintaining the erythrocyte sample in the anaerobic chamber during the NO experiment had no effect on the oxyHb levels (95.6 ± 1.0%), indicating that the conversion of oxyHb was related to the NO treatment rather than to the effects of the anaerobic chamber. This experiment was repeated twice with similar results.
Supraphysiological NO treatment results in erythrocyte lysis that diminishes P. berghei viability. To assess whether a high dose of NO rapidly kills malarial parasites ex vivo, parasite viability and red cell permeability were assessed after 10 min of incubation with 1.5 mM NO in the anaerobic chamber. Hydroethidine is a vital dye that becomes fluorescent when cleaved to ethidium by the intracellular esterases of a live cell; this conversion within cells was detected by flow cytometry. Propidium iodide is excluded when the cell membrane is intact and consequently was used to assess cell permeability. In the untreated samples (n = 3), the number of erythrocytes was 10.9 ± 0.8 x 106; the parasitemia was 10.4 ± 0.1%; and the percentage of permeable erythrocytes was 0.7 ± 0.1%. The parasitemia assessed in Giemsa-stained thin blood films was 10.5%, and the theoretical number of erythrocytes in the inoculum was 10 x 106. Immediately following incubation in PBS at 37°C for 10 min, there was no detectable change (P > 0.9) in any of the above-named three parameters (Fig. 4). Incubation of the parasite inoculum with 1.5 mM NO (n = 3) resulted in a significant (P < 0.0001) decline in the number of erythrocytes (NO, 0.93x 106 ± 0.08 x 106), but the parasitized erythrocytes were preferentially spared from NO-induced erythrocyte lysis because the parasitemia had increased significantly (P < 0.0001) to 93.9 ± 0.8%. Analysis of the forward and side scatter of the samples showed a marked increase in cell debris after NO treatment of the inocula that was not observed in the PBS controls (Fig. 4) and a significant (P < 0.0001) increase in the percentage of permeable erythrocytes (PBS, 0.7% ± 0.1%; NO, 91.0% ± 3.7%). In a replicate experiment, two out of three samples treated with 1.5 mM NO exhibited severe RBC lysis compared with PBS-treated controls at 10 min.
SIN-1 treatment has no effect on parasite viability and host survival. Because (i) macrophages during malaria produce high levels of NO and superoxide (11, 18, 19, 26, 28), which react to form peroxynitrite (4), and (ii) superoxide and peroxynitrite reportedly kill the P. falciparum parasite in vitro (16), SIN-1 treatment was used to determine whether superoxide, nitric oxide, and peroxynitrite collectively have an effect on parasite viability. Parasitized RBCs were treated with 1 mM SIN-1 for 1 h in room air and at room temperature and then injected into mice. The oxyheme concentration of the inoculum was 13.3 ± 1.5 μM. Despite supraphysiological levels of ROS, the levels of parasitemia in the mice injected with the SIN-1-treated inoculum on day 4 or day 6 were similar to those seen with the PBS control group (day 4, P = 1.0; day 6, P = 0.6) (Fig. 5A). Further, the ROS treatment had no detectable effect on the parasite's ability to elicit disease because the host survival rates for mice injected with SIN-1-treated parasites were similar to those of mice injected with control parasites (Fig. 5B). This experiment was repeated with similar results, and the SIN-1 treatment for 10 min at 37°C also did not reduce parasitemia or increase survival.
DISCUSSION
There is generally equivocal evidence that reactive oxygen species produced by phagocytes play a crucial role in controlling Plasmodium replication in vivo. However, few malarial studies take into account the role of hemoglobin as a potent scavenger of reactive oxygen species and the physiological levels of ROS that a parasitized erythrocyte might be exposed to in vivo. The malarial parasite resides in a parasitophorous vacuole within the erythrocyte and is surrounded by about 0.25 fmol of Hb within the murine erythrocyte and an additional 2.5 mM in other erythrocytes. Thus, to reach and kill an intraerythrocytic malarial parasite the level of ROS must exceed this Hb concentration. Additionally, Hb is released from the erythrocyte into blood during schizogony (erythrocyte rupture and parasite progeny release). Our preliminary data suggest that the plasma concentration of free Hb on day 6 of P. berghei infection is markedly elevated and can reach up to 100 μM just prior to death. Free hemoglobin is about 1,000-fold more efficient at scavenging NO than Hb packaged in RBCs (21), so this free Hb may also be protective by significantly quenching ROS before they can diffuse to the erythrocyte.
We observed that the Hb state in inoculum samples from day 6 animals is usually 96% oxyHb and <4% of the other forms combined. There is about 15 pg of hemoglobin or about 1 fmol of heme in a single healthy murine erythrocyte (7). The inoculum containing 7.5 million parasitized erythrocytes at about 10% parasitemia had an assayed hemoglobin concentration of 7.5 μM (30 μM total heme), translating into about 10 pg of Hb/RBC or 0.6 fmol of heme/RBC. Despite treatment with high doses of NO, no significant effect of 1.5 μM or 15 μM NO treatment upon parasitemia is detected after injection of the inoculum into mice, indicating that intraerythrocytic Plasmodium parasites are protected against ROS at up to the concentration of Hb within the erythrocytes.
Treating the parasite with 150 μM nitric oxide also has no significant effect on the parasitemia, despite the NO concentration being far in excess of the Hb concentration in the inoculum. Because these nitric oxide treatments do not kill the parasite, it is possible that some factor(s), such as the cytoskeletal diffusional barrier proposed by Liao, limits the bioavailability of the nitric oxide inside the erythrocytes by increasing the time it takes to achieve equilibrium. If the nitric oxide was able to diffuse into the erythrocyte and the Hb within the erythrocyte was protecting the parasite from NO exposure, then the NO should have converted the erythrocytic oxyHb to metHb. After 150 μM NO treatment of a sample containing parasitized erythrocytes and with about 8 μM hemoglobin, little oxyHb (5.2%) was now detectable, with the Hb being mainly in the metHb state, with some choleglobin. Thus, the ineffectiveness of the NO treatments in reducing parasite viability is not due to a failure of delivery—the 10-min treatments were sufficient to allow NO to diffuse into the RBC—but to the presence of protective mechanism(s), including hemoglobin scavenging of NO. Because there is no detectable impact on P. berghei viability after the conversion of almost all of the Hb within the erythrocyte to metHb, the parasite either uses alternate food sources or is capable of processing and storing metHb in a manner analogous to that seen with oxyHb.
Based on current research, a single parasitized erythrocyte is likely exposed to much less than 1 μM of NO in vivo (23, 34). Our data indicate that exposure to up to 150 μM NO has little or no effect on parasite viability, indicating that NO is unlikely to inhibit P. berghei replication. This conclusion is supported by the observation that the characteristics of P. berghei parasitemia and P. chabaudi parasitemia in iNOS–/– mice are similar to those seen with NOS-intact controls (12, 32, 36). One possibility is that NO toxicity is a component of a multihit sequence and that other immune mechanisms exhibit greater effects against parasites weakened by NO. This issue is difficult to address in the P. berghei model because the parasite replicates similarly in strains of mice that are resistant or susceptible to experimental severe malaria and all mice succumb to anemia secondary to hyperparasitemia, indicating the absence of a protective immune response in P. berghei-infected mice. Nevertheless, P. chabaudi-infected mice injected with heat-killed Propionibacterium acnes parasites that elicit high levels of NO production for more than a week exhibit parasitemia similar to that seen with P. acnes-injected iNOS–/– mice that lack elevated NO production, indicating that there is no detectable effect of NO toxicity with respect to rendering the parasite more susceptible to other immune mechanisms.
Besides nitric oxide, activated macrophages also produce superoxide, but at a much (10 times) lower level. Due to the much higher NO levels and the fact that both radicals come from the same source, all of the superoxide rapidly reacts with the nitric oxide, yielding 1.5 nM peroxynitrite at the macrophage surface (24). Thus, in vivo, there is little likelihood of superoxide reaching the parasite, particularly when one considers the presence of other erythrocytes and free Hb in the blood. Indeed, the observation that P mice (mice of a strain that has defective macrophage function) exhibit P. chabaudi parasitemia similar to that seen with superoxide-producing controls supports this conclusion (6) but contrasts with findings of increased P. berghei parasitemia in p91pHox–/– mice (27).
The peroxynitrite resulting from the reaction of superoxide and nitric oxide and produced by the macrophage is a highly reactive radical and at physiological pH has an extremely short half-life (24). This results in a very short diffusion distance and a near-zero concentration (<0.2 nM) at 10 μm from the macrophage (24). Thus, a sequestered erythrocyte is unlikely to be exposed to significant concentrations of peroxynitrite (<1 nM) and peroxynitrite is quenched by oxyhemoglobin to form metHb, nitrite, and oxygen. Based on the low production and its scavenging by hemoglobin, we propose that peroxynitrite does not play a significant role in parasite killing in vivo.
This conclusion is supported by our observation that 1 mM SIN-1 producing 220 μM (3.7 μM/min x 60 min) of peroxynitrite over the course of an hour had no detectable effect on P. berghei viability. This conclusion is not supported by the results of studies coculturing P. falciparum-infected erythrocytes with gamma interferon-activated macrophages; these studies have shown that nitric oxide, superoxide, and peroxynitrite reduce parasitemia and do play a role in parasite killing. However, there are concerns regarding the interpretation of these studies. First, the investigators added supraphysiological concentrations (1 mM) of SIN-1 for over 24 h to the cocultures already producing additional ROS (16), resulting in >1 mM cumulative peroxynitrite production. Second, these studies used approximate methods to assess parasitemia. Ockenhouse et al. used [3H]hypoxanthine incorporation (hypoxanthine is incorporated when the parasite replicates its DNA at trophozoite stage) to assess parasitemia (25). We subsequently reported that coculture of P. falciparum with activated monocyte-derived macrophages shortens the replication time from 48 h to 24 h; this decreased replication time results in minimal [3H]hypoxanthine incorporation despite the presence of viable parasites, because the parasites are past the stage of DNA incorporation when the radiolabel is added (31). Indeed, Fritsche et al. did not observe significant parasite killing until they added the supraphysiological levels of SIN-1 (16).
Collectively, our results indicate that P. berghei parasites remain viable even after being treated with NO (150 μM) or peroxynitrite (220 μM) at concentrations far in excess of the hemoglobin concentration (8 μM) in the inoculum. Indeed, even after we treat the parasite inoculum with a nearly saturated NO solution (1.5 mM), some parasites remain intact and are labeled with the vital dye hydroethidine. Because NO and peroxynitrite concentrations exceed the scavenging capacity of Hb in our defined ex vivo experiment by more than 1 order of magnitude, these observations indicate that P. berghei is exposed to significant concentrations of NO without exhibiting any detectable toxicity. That Plasmodium parasites likely have intrinsic protective mechanisms against NO toxicity is not surprising in light of the fact that Plasmodium parasites possess a host of antioxidant and antiradical mechanisms inherent in the pro-oxidant nature of their redox metabolism, their hemozoin granules, and their intraerythrocytic environment (3). Because all Plasmodium parasites reside within erythrocytes, the basis for our proposed mechanism of protection of P. berghei parasites from ROS, the ROS-quenching potential of hemoglobin, as well as any intrinsic protection should be conserved across all species of Plasmodium. Further, the hemoglobin content of the blood is so high compared to the NO and ROS production capabilities of the immune system that we speculate that our results can be applied to other species of parasite.
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health grants to J.A.F. (HL40696), M.I. (R24 HL643952), and H.V.D.H. (AI40667).
The authors have no conflicting financial interests.
REFERENCES
1. Alayash, A. I. 2001. Oxidative mechanisms of hemoglobin-based blood substitutes. Artif. Cells Blood Substit. Immobil. Biotechnol. 29:415-425.
2. Alayash, A. I. 2004. Oxygen therapeutics: can we tame haemoglobin Nat. Rev. Drug Discov. 3:152-159.
3. Becker, K., L. Tilley, J. L. Vennerstrom, D. Roberts, S. Rogerson, and H. Ginsburg. 2004. Oxidative stress in malaria parasite-infected erythrocytes: host-parasite interactions. Int. J. Parasitol. 34:163-189.
4. Bogdan, C. 2001. Nitric oxide and the immune response. Nat. Immunol. 2:907-916.
5. Brunet, L. R. 2001. Nitric oxide in parasitic infections. Int. Immunopharmacol. 1:1457-1467.
6. Cavacini, L. A., M. Guidotti, L. A. Parke, J. Melancon-Kaplan, and W. P. Weidanz. 1989. Reassessment of the role of splenic leukocyte oxidative activity and macrophage activation in expression of immunity to malaria. Infect. Immun. 57:3677-3682.
7. Chen, D., and D. K. Kaul. 1994. Rheologic and hemodynamic characteristics of red cells of mouse, rat and human. Biorheology 31:103-113.
8. Clark, I. A., and W. B. Cowden. 2003. The pathophysiology of falciparum malaria. Pharmacol. Ther. 99:221-260.
9. Cripps, C. M. 1968. Rapid method for the estimation of plasma haemoglobin levels. J. Clin. Pathol. 21:110-112.
10. Elloso, M. M., H. C. van der Heyde, J. A. vande Waa, D. D. Manning, and W. P. Weidanz. 1994. Inhibition of Plasmodium falciparum in vitro by human gamma delta T cells. J. Immunol. 153:1187-1194.
11. Erel, O., H. Vural, N. Aksoy, G. Aslan, and M. Ulukanligil. 2001. Oxidative stress of platelets and thrombocytopenia in patients with vivax malaria. Clin. Biochem. 34:341-344.
12. Favre, N., B. Ryffel, and W. Rudin. 1999. Parasite killing in murine malaria does not require nitric oxide production. Parasitology 118(Part 2):139-143.
13. Feelisch, M. and J. S. Stamler. 1996. Methods in nitric oxide research. John Wiley and Sons, Chichester, England.
14. Finley, R., J. Weintraub, J. A. Louis, H. D. Engers, R. Zubler, and P. H. Lambert. 1983. Prevention of cerebral malaria by adoptive transfer of malaria-specific cultured T cells into mice infected with Plasmodium berghei. J. Immunol. 131:1522-1526.
15. Finley, R. W., L. J. Mackey, and P. H. Lambert. 1982. Virulent P. berghei malaria: prolonged survival and decreased cerebral pathology in cell-dependent nude mice. J. Immunol. 129:2213-2218.
16. Fritsche, G., C. Larcher, H. Schennach, and G. Weiss. 2001. Regulatory interactions between iron and nitric oxide metabolism for immune defense against Plasmodium falciparum infection. J. Infect. Dis. 183:1388-1394.
17. Hogg, N., V. M. Darley-Usmar, M. T. Wilson, and S. Moncada. 1992. Production of hydroxyl radicals from the simultaneous generation of superoxide and nitric oxide. Biochem. J. 281(Part 2):419-424.
18. Jaramillo, M., I. Plante, N. Ouellet, K. Vandal, P. A. Tessier, and M. Olivier. 2004. Hemozoin-inducible proinflammatory events in vivo: potential role in malaria infection. J. Immunol. 172:3101-3110.
19. Krishnegowda, G., A. M. Hajjar, J. Zhu, E. J. Douglass, S. Uematsu, S. Akira, A. S. Woods, and D. C. Gowda. 2005. Induction of proinflammatory responses in macrophages by the glycosylphosphatidylinositols (GPIs) of Plasmodium falciparum: cell signaling receptors, GPI structural requirement, and regulation of GPI activity. J. Biol. Chem. 280:8606-8616.
20. Krogstad, D. J. 1996. Malaria as a reemerging disease. Epidemiol. Rev. 18:77-89.
21. Liu, X., M. J. Miller, M. S. Joshi, H. Sadowska-Krowicka, D. A. Clark, and J. R. Lancaster, Jr. 1998. Diffusion-limited reaction of free nitric oxide with erythrocytes. J. Biol. Chem. 273:18709-18713.
22. Murphy, M. E., and E. Noack. 1994. Nitric oxide assay using hemoglobin method. Methods Enzymol. 233:240-250.
23. Nalwaya, N., and W. M. Deen. 2004. Analysis of the effects of nitric oxide and oxygen on nitric oxide production by macrophages. J. Theor. Biol. 226:409-419.
24. Nalwaya, N., and W. M. Deen. 2004. Peroxynitrite exposure of cells cocultured with macrophages. Ann. Biomed. Eng. 32:664-676.
25. Ockenhouse, C. F., S. Schulman, and H. L. Shear. 1984. Induction of crisis forms in the human malaria parasite Plasmodium falciparum by gamma-interferon-activated, monocyte-derived macrophages. J. Immunol. 133:1601-1608.
26. Postma, N. S., E. C. Mommers, W. M. Eling, and J. Zuidema. 1996. Oxidative stress in malaria; implications for prevention and therapy. Pharm. World Sci. 18:121-129.
27. Sanni, L. A., S. Fu, R. T. Dean, G. Bloomfield, R. Stocker, G. Chaudhri, M. C. Dinauer, and N. H. Hunt. 1999. Are reactive oxygen species involved in the pathogenesis of murine cerebral malaria J. Infect. Dis. 179:217-222.
28. Schofield, L., and F. Hackett. 1993. Signal transduction in host cells by a glycosylphosphatidylinositol toxin of malaria parasites. J. Exp. Med. 177:145-153.
29. Stevenson, M. M., and E. M. Riley. 2004. Innate immunity to malaria. Nat. Rev. Immunol. 4:169-180.
30. Taylor-Robinson, A. W., R. S. Phillips, A. Severn, S. Moncada, and F. Y. Liew. 1993. The role of TH1 and TH2 cells in a rodent malaria infection. Science 260:1931-1934.
31. van der Heyde, H. C., M. M. Elloso, J. vande Waa, K. Schell, and W. P. Weidanz. 1995. Use of hydroethidine and flow cytometry to assess the effects of leukocytes on the malarial parasite Plasmodium falciparum. Clin. Diagn. Lab. Immunol. 2:417-425.
32. van der Heyde, H. C., Y. Gu, Q. Zhang, G. Sun, and M. B. Grisham. 2000. Nitric oxide is neither necessary nor sufficient for resolution of Plasmodium chabaudi malaria in mice. J. Immunol. 165:3317-3323.
33. Vaughn, M. W., K. T. Huang, L. Kuo, and J. C. Liao. 2000. Erythrocytes possess an intrinsic barrier to nitric oxide consumption. J. Biol. Chem. 275:2342-2348.
34. Vaughn, M. W., L. Kuo, and J. C. Liao. 1998. Estimation of nitric oxide production and reaction rates in tissue by use of a mathematical model. Am. J. Physiol. 274:H2163-H2176.
35. Winterbourn, C. C. 1990. Oxidative reactions of hemoglobin. Methods Enzymol. 186:265-272.
36. Yoneto, T., T. Yoshimoto, C. R. Wang, Y. Takahama, M. Tsuji, S. Waki, and H. Nariuchi. 1999. Gamma interferon production is critical for protective immunity to infection with blood-stage Plasmodium berghei XAT but neither NO production nor NK cell activation is critical. Infect. Immun. 67:2349-2356.(Peter Sobolewski, Irene G)
ABSTRACT
Reactive oxygen species (ROS) are widely believed to kill malarial parasites. C57BL/6 mice injected with P. berghei inocula incubated with supraphysiological doses of NO (150 μM) or with peroxynitrite (220 μM), however, exhibited parasitemia similar to that seen with those given control inocula, and there was no difference in disease development. Only treatment of inocula with NO doses nearing saturation (1.2 mM) resulted in no detectable parasitemia in the recipients; flow cytometric analysis with a vital dye (hydroethidine) indicated that 1.5 mM NO lysed the erythrocytes rather than killing the parasites. The hemoglobin level in the inocula was about 8 μM; the hemoglobin was mainly oxyhemoglobin (oxyHb) (96%), which was converted to methemoglobin (>95%) after treatment with 150 μM NO. The concentrations of 150 μM of NO and 220 μM of peroxynitrite were far in excess of the hemoglobin concentration (8 μM), and yet no parasite killing was detected. We therefore conclude that hemoglobin protects Plasmodium parasites from ROS, but the parasite likely possesses intrinsic defense mechanisms against ROS.
INTRODUCTION
Malaria, a reemerging disease (20) caused by the parasite of the genus Plasmodium, remains refractory to the development of a vaccine in part due to incomplete understanding of the mechanism(s) underlying parasite killing by the immune system. It is generally accepted that reactive oxygen species (ROS), including nitric oxide (NO), superoxide, and peroxynitrite, kill intraerythrocytic malarial parasites (5, 8, 29). The most cited mechanism for parasite killing is that acute Plasmodium infection induces gamma interferon-producing Th1 cells, which in turn activate macrophages to secrete parasiticidal NO and ROS (29, 30).
We propose, however, that the blood stage Plasmodium parasite is virtually immune to the cytotoxic effects of NO and ROS as a consequence of hemoglobin (Hb) NO scavenging and ROS suppression within red blood cells (RBCs). The Plasmodium parasite is surrounded by hemoglobin through most of its asexual blood cycle, because it resides within a parasitophorous vacuole inside erythrocytes. Plasmodium falciparum parasites rupture erythrocytes, releasing progeny merozoites, which invade new RBCs after completing their 48-hour blood stage cycle. This extracellular excursion constitutes a brief period in which the parasite is in principle vulnerable to higher ROS concentrations induced by the infection. However, the disruption of the RBC membrane inevitably releases molecular hemoglobin into the circulation, enhancing ROS scavenging; thus, both inside and outside the red cells the parasite is protected from ROS because ROS are scavenged by hemoglobin.
Malaria, therefore, is fundamentally different from most infections, because the parasite is surrounded by hemoglobin and can evade the ROS-based protective mechanism as a consequence of ROS quenching by Hb (2), an antioxidant mechanism that has been overlooked. Although Hb's heme group can undergo redox transitions to higher oxidation states and it can auto-oxidize naturally to form methemoglobin (metHb) and superoxide ions, this oxidation process is controlled within RBCs by the metHb reductase system (1). The biochemical basis for the in vivo hemoglobin redox reactions has been extensively reviewed by Alayash (2).
To test the hypothesis that the Plasmodium parasite evades the ROS-based protective mechanisms because the parasite is surrounded by hemoglobin, we analyzed NO and ROS effects on Plasmodium berghei ex vivo, with particular attention to the molecular state of hemoglobin. Although there are differences between human and experimental malaria which limit extrapolation to the human condition, ex vivo treatments of P. berghei can assess the effect of NO and ROS on common features of parasite viability and replication as well as its ability to elicit disease.
MATERIALS AND METHODS
Evaluation of NO solution bioactivity. (i) Nitric oxide stock preparation. Phosphate-buffered saline (PBS) solution (Invitrogen, Carlsbad, CA) (pH 7.4) was degassed under vacuum overnight and placed in line between a 2 M NaOH solution and a 1 M KMnO4 solution. Prepurified grade 4.8 (99.998% pure) nitrogen was bubbled through the system for 30 min followed by UHP nitric oxide (Airgas, San Diego, CA) (99.5% pure) for 5 min (13). A saturated solution of NO has a concentration of 1.8 mM at room temperature and 1 atm (13), which we verified using the amperometric inno-T nitric oxide system (Innovative Instruments, Inc.). NO stock solutions were prepared in an anaerobic chamber filled with grade 4.8 nitrogen gas, 2 μl or 5 μl (1:10,000 and 1:4,000 dilution, respectively) was injected into a stirred 20 ml sample of PBS, and the resultant NO concentrations were monitored as a function of time and were 176.9 ± 6.6 nM for the 10,000 dilution and 435.6 ± 2.3 nM for the 4,000 dilution.
(ii) Hemoglobin solution preparation and NO concentration analysis. Hemoglobin solutions were prepared by lysing freshly isolated murine RBCs. Blood was obtained from anesthetized C57BL/6 mice (Jackson Laboratories) and centrifuged at 500 x g for 10 min to pellet the erythrocytes. The supernatant was aspirated, and then the cells were lysed with 0.9 ml distilled water. The solution was later osmotically balanced with the addition of 0.1 ml 10x PBS. The oxyhemoglobin (oxyHb) and methemoglobin concentration of this stock was determined using the Winterbourn spectrophotometric method (35). The oxyhemoglobin assay for NO (13, 22) was used to verify the effective dose of NO treatments. Hemoglobin solutions of 1.2 mM, 600 μM, 60 μM, and 20 μM were prepared and were treated three times using Hamilton syringes with 0.7 ml, 125 μl, 12.5 μl, and 1.25 μl of saturated NO solutions, respectively. Following each injection, the change in oxyHb level was determined by monitoring the change in absorption at 577 nm with a μQuant spectrophotometer (BioTek, Winooski, VT).
Infection of mice with ROS-treated P. berghei. Plasmodium berghei ANKA was stored as a frozen stabilate. The stabilate was thawed and then injected into a source mouse to generate an inoculum for the experiments. Blood from the source mouse was diluted with PBS to prepare aliquots of the inoculum. In the first series of experiments, the blood was diluted 1:6 in PBS upon collection from the source mouse. The number of RBC per milliliter was determined by using a hemocytometer. This value and the percent parasitemia were used to calculate the final dilution necessary to obtain 7.5 x 106 parasitized erythrocytes (pRBCs) in 1.5 ml PBS (5 x 106 pRBC/ml), and this dilution (100-fold) was then achieved. In subsequent experiments, the blood cells were kept in a 1:6 dilution of PBS and then diluted to the final concentration just before they were exposed to either NO or SIN-1 (3-morpholino-sydnonimine) as described below. A 200-μl volume of the NO- or SIN-1-exposed inoculum (1 x 106 pRBC) was injected intraperitoneally (i.p.) into groups of three to five C57BL/6J mice for each treatment group. The mice were obtained from Jackson Laboratories (Bar Harbor, ME) at 6 to 8 weeks of age and were injected with parasitized RBCs (pRBCs) when they were 8 to 10 weeks of age. The animals were housed in microisolator cages and provided food ad libitum. Parasitemia was assessed in each group of mice by counting the number of infected erythrocytes in Giemsa-stained thin blood films. The effects of NO treatments on the survival of severe malaria were also assessed. The P. berghei model of malaria is a very reproducible model of cerebral malaria in which virtually all C57BL/6J mice develop neurologic impairment and obtundation and then succumb to infection between days 6 and 8 of infection (14, 15). The Institutional Animal Care and Use Committee committee at the La Jolla Bioengineering Institute approved all procedures.
(i) Nitric oxide treatment. Experiments were carried out in a water-jacketed reaction chamber (WPI), preheated to 37°C, which was placed inside a custom anaerobic chamber that was degassed with grade 4.8 (99.998% pure) nitrogen for 1 h. Inoculum samples containing 7.5 million pRBCs were pipetted into the reaction chamber and treated with a bolus of nitric oxide stock solution (1,250 μl, 1,000 μl, 125 μl, 12.5 μl, or 1.25 μl) delivered via a gas-tight Hamilton syringe. The final sample volume was 1.5 ml, contained 7.5 million pRBC, and had an NO concentration of 1.5 mM, 1.2 mM, 150 μM, 15 μM, or 1.5 μM. The chamber was capped to minimize NO escape from the solution into the gas headspace, and the solution was maintained at 37°C for 10 min. The control treatment consisted of pipetting 1,250 μl of degassed PBS into the 250 μl inoculum sample containing 7.5 million pRBC. At the end of the treatment period, the treated inoculum was aspirated from the sample chamber via syringe, and 200 μl of the sample (1 million pRBC per sample) was injected i.p. into mice. The oxyhemoglobin content of inoculum was determined by lysing the erythrocytes and analyzing the lysate by use of the Cripps method (9).
(ii) NO bioactivity assessment. The inoculum was treated with 1.5 mM NO, 150 μM NO, or PBS, as described above. Instead of splitting the inoculum into samples and injecting the samples into mice, it was then centrifuged and lysed, as described above, and the Winterbourn spectrophotometric assay was used to determine oxyHb, metHb, and choleglobin levels.
(iii) SIN-1 treatment. A SIN-1 stock solution (250 mM) was prepared by dissolving 50 mg SIN-1 chloride (Alexis, San Diego, CA) in PBS and immediately freezing the solution. SIN-1 treatments consisted of adding 8 μl of SIN-1 stock to 1 ml PBS and mixing this resultant 2 mM SIN-1 solution with 1 ml of inoculum containing 10 million pRBC in PBS, yielding 2 ml of 1 mM SIN-1 solution with 5 million pRBC/ml and production rates of superoxide and nitric oxide of 7.0 μM/min and 3.7 μM/min, respectively (17). The treated inoculum was maintained in room air for either 1 h at room temperature or 10 min at 37°C. Following treatment, the inoculum was centrifuged at 500 x g for 10 min and resuspended in 2 ml of PBS, and 200 μl samples (1 million pRBC) were injected i.p. into mice as described above.
Flow cytometric assessment of parasite viability and permeability of RBCs. Inoculum samples were labeled with the vital dye hydroethidine immediately after NO treatment or following a 4-h incubation in fetal bovine serum at 37°C by the use of a previously described technique (10). Hydroethidine (Molecular Probes) at 10 mg/ml in dimethyl sulfoxide (Invitrogen Corporation, Carlsbad, CA) was diluted 1:600 in PBS. Parasite inocula (10 million) treated with 1.5 mM NO or PBS were washed once in PBS, resuspended in 0.5 ml of the hydroethidine solution, and incubated for 20 min at 37°C in the dark. Following the incubation, all hydroethidine-stained samples were washed twice in PBS and resuspended in 1 ml of PBS. Propidium iodide was added to replicates of the treated inocula to assess the number of erythrocytes that were permeable. Spherotech beads were added to all samples prior to being processed using a FACSCalibur system (Beckton Dickinson, San Diego, CA) to assess cell numbers. These beads are distinguished from cells on the basis of their unique forward and side scatter and fluorescence on all three channels. The acquisition of data was performed using the CellQuest (Becton-Dickinson) program on 10,000 cells identified by characteristic forward and side scatter. The Attractors (Becton-Dickinson) program was used to analyze the data. Cell numbers were determined using the following formula:
Statistical analysis. Analysis of variance with the Statview program (SAS Institute, Cary, NC) and Fisher's post hoc test was performed to statistically compare all parasitemia with a P value cutoff of 0.05. The means and standard deviations of the results are reported in the text and figures.
RESULTS
Verification of nitric oxide stock concentration. The decay of NO concentration with time was monitored with a NO amperometric method, showing that our stock was saturated with NO and that dilutions of the stock provided reproducible treatments of NO. The recording of NO levels in the solution diluted 1:4,000 showed that the exponential decay constant for NO in the anaerobic chamber was 20.9 ± 0.6 min. It was not possible to eliminate the headspace above the solution, so the most-rapid loss of NO was due to NO diffusion from the solution into the atmosphere. This NO diffusion was minimized in the capped reaction chamber used in the treatment of parasite inocula, so we estimate that the NO decay time constant for parasite treatments is at least twice that measured (41.8 ± 1.2 min). Either time constant indicates that most of the NO was present during the 10-min treatment period in the anaerobic chamber.
The amount of NO delivered by our treatments and by the treatment variations was determined using the oxyhemoglobin NO assay because this is the NO experienced by the RBC-based parasite, and concentrations were beyond the operating range of our NO electrode. The Hb solutions were treated with the selected NO doses, and then the amount of bioactive NO actually delivered was assessed by spectroscopically monitoring the reduction in oxyHb (13, 22). Before treatment, most (90.7 ± 2.5%) of the Hb in the test solution was oxyHb. Measurement of oxyHb concentration following treatment showed that the selected doses of NO were similar to the actual concentrations for bioactive NO (Fig. 1B).
Almost-saturated solutions of nitric oxide are needed to impair the viability of P. berghei parasites. There was no detectable effect of NO on parasitemia or development of disease in groups of mice injected i.p. with 1 x 106 P. berghei pRBCs treated with either 15 μM or 1.5 μM NO compared to the control group results (Fig. 2A). In this experiment, blood was immediately diluted (1:600) to obtain the parasite inoculum (7.5 x106 pRBC/1.5 ml) and was kept diluted in PBS for about 2 h before the injection of 1 x 106 pRBCs into the mice. Mice injected with parasitized RBCs treated with 150 μM NO showed a modest but not significant decrease in parasitemia at days 4 and 6 (P = 0.17 and P = 0.63, respectively) and slightly delayed development of experimental cerebral malaria (Fig. 2B). Thus, there appeared to be an increase in survival in groups of mice injected with parasitized RBCs treated with 150 μM NO (Fig. 2B). Only mice infected with parasitized RBCs treated with 1.5 mM NO exhibited no viable parasites and consequently no development of disease. This experiment was repeated twice with similar results.
We assessed the amount of Hb in the parasitized erythrocyte sample and assessed the state of the Hb within the erythrocytes prior to our NO treatment. The oxyheme concentration of the samples assessed by the Cripps spectrophotometric method was 29.3 ± 1.4 μM. Thus, the samples treated with 150 μM and 1.5 mM NO had sufficient NO in principle to overcome the Hb quenching of NO.
We repeated the above-described experiment except that instead of diluting the blood immediately to its final concentration (7.5 x106 pRBC/1.5 ml) for NO treatment, the blood was kept at a 1:6 dilution and then diluted further (1:100) to the final inoculum concentration just before exposure to NO. Mice were injected i.p. with 1 x 106 pRBCs from the inoculum immediately after it was exposed to NO. We performed this assay because hydroethidine results suggested that prolonged incubation in PBS alone lowered parasite viability compared with the results seen with parasite samples kept in a 1:6 dilution and then diluted immediately before use; this suggests that the lack of nutrients may have affected parasite viability. Under the new condition of parasite storage prior to NO treatment, no change in parasite viability or disease-eliciting capability was now detected at the 150 μM NO concentration (Fig. 2C). This observation indicated that a weakening of the parasite by prolonged (1 to 2 h) incubation in the absence of serum had occurred, rendering it artificially more susceptible to NO killing. The 1.2 mM NO treatment resulted in no viable parasites, and no infected mice developed disease (Fig. 2D). The oxyheme concentration in the inoculum samples was 45.6 ± 1.3 μM, indicating that the 150 μM and 1.2 mM NO treatments should have overcome the NO-quenching capacity of the oxyHb. We repeated this experiment twice with similar results. In all cases, the concentrations used for the NO treatments (1.2 mM and 150 μM) were far in excess of the oxyhemoglobin concentration of the samples.
Nitric oxide is bioactive following diffusion into erythrocytes. Because no parasite killing was detected even when the NO treatment concentration (150 μM) far exceeded the sample hemoglobin concentration (30 to 45 μM), it is possible that sufficient NO did not diffuse into the RBC during the course of the 10-min treatments. In fact, Liao and coworkers (33) have proposed that diffusional barriers slow the kinetics of NO consumption by RBCs, thereby explaining why NO consumption by free Hb is about 1,000-fold greater than that seen with the same amount of Hb packaged in RBCs. To verify that on the 10-min time scale of our treatments the diffusional barriers did not markedly inhibit NO bioavailability within the erythrocyte and that the parasite was exposed to NO, we determined the Hb oxidation state before and immediately after NO treatment of the experimental samples. Before NO treatment and placement in the anaerobic chamber, the samples were mainly oxyHb (96.1 ± 1.1%), with little metHb (3.9 ± 1.1%) or choleglobin (0.0 ± 0.0%), a denatured form of Hb. Following NO treatment of the parasitized erythrocytes, most of the oxyHb (95%) was converted into either metHb or choleglobin (Fig. 3). Maintaining the erythrocyte sample in the anaerobic chamber during the NO experiment had no effect on the oxyHb levels (95.6 ± 1.0%), indicating that the conversion of oxyHb was related to the NO treatment rather than to the effects of the anaerobic chamber. This experiment was repeated twice with similar results.
Supraphysiological NO treatment results in erythrocyte lysis that diminishes P. berghei viability. To assess whether a high dose of NO rapidly kills malarial parasites ex vivo, parasite viability and red cell permeability were assessed after 10 min of incubation with 1.5 mM NO in the anaerobic chamber. Hydroethidine is a vital dye that becomes fluorescent when cleaved to ethidium by the intracellular esterases of a live cell; this conversion within cells was detected by flow cytometry. Propidium iodide is excluded when the cell membrane is intact and consequently was used to assess cell permeability. In the untreated samples (n = 3), the number of erythrocytes was 10.9 ± 0.8 x 106; the parasitemia was 10.4 ± 0.1%; and the percentage of permeable erythrocytes was 0.7 ± 0.1%. The parasitemia assessed in Giemsa-stained thin blood films was 10.5%, and the theoretical number of erythrocytes in the inoculum was 10 x 106. Immediately following incubation in PBS at 37°C for 10 min, there was no detectable change (P > 0.9) in any of the above-named three parameters (Fig. 4). Incubation of the parasite inoculum with 1.5 mM NO (n = 3) resulted in a significant (P < 0.0001) decline in the number of erythrocytes (NO, 0.93x 106 ± 0.08 x 106), but the parasitized erythrocytes were preferentially spared from NO-induced erythrocyte lysis because the parasitemia had increased significantly (P < 0.0001) to 93.9 ± 0.8%. Analysis of the forward and side scatter of the samples showed a marked increase in cell debris after NO treatment of the inocula that was not observed in the PBS controls (Fig. 4) and a significant (P < 0.0001) increase in the percentage of permeable erythrocytes (PBS, 0.7% ± 0.1%; NO, 91.0% ± 3.7%). In a replicate experiment, two out of three samples treated with 1.5 mM NO exhibited severe RBC lysis compared with PBS-treated controls at 10 min.
SIN-1 treatment has no effect on parasite viability and host survival. Because (i) macrophages during malaria produce high levels of NO and superoxide (11, 18, 19, 26, 28), which react to form peroxynitrite (4), and (ii) superoxide and peroxynitrite reportedly kill the P. falciparum parasite in vitro (16), SIN-1 treatment was used to determine whether superoxide, nitric oxide, and peroxynitrite collectively have an effect on parasite viability. Parasitized RBCs were treated with 1 mM SIN-1 for 1 h in room air and at room temperature and then injected into mice. The oxyheme concentration of the inoculum was 13.3 ± 1.5 μM. Despite supraphysiological levels of ROS, the levels of parasitemia in the mice injected with the SIN-1-treated inoculum on day 4 or day 6 were similar to those seen with the PBS control group (day 4, P = 1.0; day 6, P = 0.6) (Fig. 5A). Further, the ROS treatment had no detectable effect on the parasite's ability to elicit disease because the host survival rates for mice injected with SIN-1-treated parasites were similar to those of mice injected with control parasites (Fig. 5B). This experiment was repeated with similar results, and the SIN-1 treatment for 10 min at 37°C also did not reduce parasitemia or increase survival.
DISCUSSION
There is generally equivocal evidence that reactive oxygen species produced by phagocytes play a crucial role in controlling Plasmodium replication in vivo. However, few malarial studies take into account the role of hemoglobin as a potent scavenger of reactive oxygen species and the physiological levels of ROS that a parasitized erythrocyte might be exposed to in vivo. The malarial parasite resides in a parasitophorous vacuole within the erythrocyte and is surrounded by about 0.25 fmol of Hb within the murine erythrocyte and an additional 2.5 mM in other erythrocytes. Thus, to reach and kill an intraerythrocytic malarial parasite the level of ROS must exceed this Hb concentration. Additionally, Hb is released from the erythrocyte into blood during schizogony (erythrocyte rupture and parasite progeny release). Our preliminary data suggest that the plasma concentration of free Hb on day 6 of P. berghei infection is markedly elevated and can reach up to 100 μM just prior to death. Free hemoglobin is about 1,000-fold more efficient at scavenging NO than Hb packaged in RBCs (21), so this free Hb may also be protective by significantly quenching ROS before they can diffuse to the erythrocyte.
We observed that the Hb state in inoculum samples from day 6 animals is usually 96% oxyHb and <4% of the other forms combined. There is about 15 pg of hemoglobin or about 1 fmol of heme in a single healthy murine erythrocyte (7). The inoculum containing 7.5 million parasitized erythrocytes at about 10% parasitemia had an assayed hemoglobin concentration of 7.5 μM (30 μM total heme), translating into about 10 pg of Hb/RBC or 0.6 fmol of heme/RBC. Despite treatment with high doses of NO, no significant effect of 1.5 μM or 15 μM NO treatment upon parasitemia is detected after injection of the inoculum into mice, indicating that intraerythrocytic Plasmodium parasites are protected against ROS at up to the concentration of Hb within the erythrocytes.
Treating the parasite with 150 μM nitric oxide also has no significant effect on the parasitemia, despite the NO concentration being far in excess of the Hb concentration in the inoculum. Because these nitric oxide treatments do not kill the parasite, it is possible that some factor(s), such as the cytoskeletal diffusional barrier proposed by Liao, limits the bioavailability of the nitric oxide inside the erythrocytes by increasing the time it takes to achieve equilibrium. If the nitric oxide was able to diffuse into the erythrocyte and the Hb within the erythrocyte was protecting the parasite from NO exposure, then the NO should have converted the erythrocytic oxyHb to metHb. After 150 μM NO treatment of a sample containing parasitized erythrocytes and with about 8 μM hemoglobin, little oxyHb (5.2%) was now detectable, with the Hb being mainly in the metHb state, with some choleglobin. Thus, the ineffectiveness of the NO treatments in reducing parasite viability is not due to a failure of delivery—the 10-min treatments were sufficient to allow NO to diffuse into the RBC—but to the presence of protective mechanism(s), including hemoglobin scavenging of NO. Because there is no detectable impact on P. berghei viability after the conversion of almost all of the Hb within the erythrocyte to metHb, the parasite either uses alternate food sources or is capable of processing and storing metHb in a manner analogous to that seen with oxyHb.
Based on current research, a single parasitized erythrocyte is likely exposed to much less than 1 μM of NO in vivo (23, 34). Our data indicate that exposure to up to 150 μM NO has little or no effect on parasite viability, indicating that NO is unlikely to inhibit P. berghei replication. This conclusion is supported by the observation that the characteristics of P. berghei parasitemia and P. chabaudi parasitemia in iNOS–/– mice are similar to those seen with NOS-intact controls (12, 32, 36). One possibility is that NO toxicity is a component of a multihit sequence and that other immune mechanisms exhibit greater effects against parasites weakened by NO. This issue is difficult to address in the P. berghei model because the parasite replicates similarly in strains of mice that are resistant or susceptible to experimental severe malaria and all mice succumb to anemia secondary to hyperparasitemia, indicating the absence of a protective immune response in P. berghei-infected mice. Nevertheless, P. chabaudi-infected mice injected with heat-killed Propionibacterium acnes parasites that elicit high levels of NO production for more than a week exhibit parasitemia similar to that seen with P. acnes-injected iNOS–/– mice that lack elevated NO production, indicating that there is no detectable effect of NO toxicity with respect to rendering the parasite more susceptible to other immune mechanisms.
Besides nitric oxide, activated macrophages also produce superoxide, but at a much (10 times) lower level. Due to the much higher NO levels and the fact that both radicals come from the same source, all of the superoxide rapidly reacts with the nitric oxide, yielding 1.5 nM peroxynitrite at the macrophage surface (24). Thus, in vivo, there is little likelihood of superoxide reaching the parasite, particularly when one considers the presence of other erythrocytes and free Hb in the blood. Indeed, the observation that P mice (mice of a strain that has defective macrophage function) exhibit P. chabaudi parasitemia similar to that seen with superoxide-producing controls supports this conclusion (6) but contrasts with findings of increased P. berghei parasitemia in p91pHox–/– mice (27).
The peroxynitrite resulting from the reaction of superoxide and nitric oxide and produced by the macrophage is a highly reactive radical and at physiological pH has an extremely short half-life (24). This results in a very short diffusion distance and a near-zero concentration (<0.2 nM) at 10 μm from the macrophage (24). Thus, a sequestered erythrocyte is unlikely to be exposed to significant concentrations of peroxynitrite (<1 nM) and peroxynitrite is quenched by oxyhemoglobin to form metHb, nitrite, and oxygen. Based on the low production and its scavenging by hemoglobin, we propose that peroxynitrite does not play a significant role in parasite killing in vivo.
This conclusion is supported by our observation that 1 mM SIN-1 producing 220 μM (3.7 μM/min x 60 min) of peroxynitrite over the course of an hour had no detectable effect on P. berghei viability. This conclusion is not supported by the results of studies coculturing P. falciparum-infected erythrocytes with gamma interferon-activated macrophages; these studies have shown that nitric oxide, superoxide, and peroxynitrite reduce parasitemia and do play a role in parasite killing. However, there are concerns regarding the interpretation of these studies. First, the investigators added supraphysiological concentrations (1 mM) of SIN-1 for over 24 h to the cocultures already producing additional ROS (16), resulting in >1 mM cumulative peroxynitrite production. Second, these studies used approximate methods to assess parasitemia. Ockenhouse et al. used [3H]hypoxanthine incorporation (hypoxanthine is incorporated when the parasite replicates its DNA at trophozoite stage) to assess parasitemia (25). We subsequently reported that coculture of P. falciparum with activated monocyte-derived macrophages shortens the replication time from 48 h to 24 h; this decreased replication time results in minimal [3H]hypoxanthine incorporation despite the presence of viable parasites, because the parasites are past the stage of DNA incorporation when the radiolabel is added (31). Indeed, Fritsche et al. did not observe significant parasite killing until they added the supraphysiological levels of SIN-1 (16).
Collectively, our results indicate that P. berghei parasites remain viable even after being treated with NO (150 μM) or peroxynitrite (220 μM) at concentrations far in excess of the hemoglobin concentration (8 μM) in the inoculum. Indeed, even after we treat the parasite inoculum with a nearly saturated NO solution (1.5 mM), some parasites remain intact and are labeled with the vital dye hydroethidine. Because NO and peroxynitrite concentrations exceed the scavenging capacity of Hb in our defined ex vivo experiment by more than 1 order of magnitude, these observations indicate that P. berghei is exposed to significant concentrations of NO without exhibiting any detectable toxicity. That Plasmodium parasites likely have intrinsic protective mechanisms against NO toxicity is not surprising in light of the fact that Plasmodium parasites possess a host of antioxidant and antiradical mechanisms inherent in the pro-oxidant nature of their redox metabolism, their hemozoin granules, and their intraerythrocytic environment (3). Because all Plasmodium parasites reside within erythrocytes, the basis for our proposed mechanism of protection of P. berghei parasites from ROS, the ROS-quenching potential of hemoglobin, as well as any intrinsic protection should be conserved across all species of Plasmodium. Further, the hemoglobin content of the blood is so high compared to the NO and ROS production capabilities of the immune system that we speculate that our results can be applied to other species of parasite.
ACKNOWLEDGMENTS
This work was supported by National Institutes of Health grants to J.A.F. (HL40696), M.I. (R24 HL643952), and H.V.D.H. (AI40667).
The authors have no conflicting financial interests.
REFERENCES
1. Alayash, A. I. 2001. Oxidative mechanisms of hemoglobin-based blood substitutes. Artif. Cells Blood Substit. Immobil. Biotechnol. 29:415-425.
2. Alayash, A. I. 2004. Oxygen therapeutics: can we tame haemoglobin Nat. Rev. Drug Discov. 3:152-159.
3. Becker, K., L. Tilley, J. L. Vennerstrom, D. Roberts, S. Rogerson, and H. Ginsburg. 2004. Oxidative stress in malaria parasite-infected erythrocytes: host-parasite interactions. Int. J. Parasitol. 34:163-189.
4. Bogdan, C. 2001. Nitric oxide and the immune response. Nat. Immunol. 2:907-916.
5. Brunet, L. R. 2001. Nitric oxide in parasitic infections. Int. Immunopharmacol. 1:1457-1467.
6. Cavacini, L. A., M. Guidotti, L. A. Parke, J. Melancon-Kaplan, and W. P. Weidanz. 1989. Reassessment of the role of splenic leukocyte oxidative activity and macrophage activation in expression of immunity to malaria. Infect. Immun. 57:3677-3682.
7. Chen, D., and D. K. Kaul. 1994. Rheologic and hemodynamic characteristics of red cells of mouse, rat and human. Biorheology 31:103-113.
8. Clark, I. A., and W. B. Cowden. 2003. The pathophysiology of falciparum malaria. Pharmacol. Ther. 99:221-260.
9. Cripps, C. M. 1968. Rapid method for the estimation of plasma haemoglobin levels. J. Clin. Pathol. 21:110-112.
10. Elloso, M. M., H. C. van der Heyde, J. A. vande Waa, D. D. Manning, and W. P. Weidanz. 1994. Inhibition of Plasmodium falciparum in vitro by human gamma delta T cells. J. Immunol. 153:1187-1194.
11. Erel, O., H. Vural, N. Aksoy, G. Aslan, and M. Ulukanligil. 2001. Oxidative stress of platelets and thrombocytopenia in patients with vivax malaria. Clin. Biochem. 34:341-344.
12. Favre, N., B. Ryffel, and W. Rudin. 1999. Parasite killing in murine malaria does not require nitric oxide production. Parasitology 118(Part 2):139-143.
13. Feelisch, M. and J. S. Stamler. 1996. Methods in nitric oxide research. John Wiley and Sons, Chichester, England.
14. Finley, R., J. Weintraub, J. A. Louis, H. D. Engers, R. Zubler, and P. H. Lambert. 1983. Prevention of cerebral malaria by adoptive transfer of malaria-specific cultured T cells into mice infected with Plasmodium berghei. J. Immunol. 131:1522-1526.
15. Finley, R. W., L. J. Mackey, and P. H. Lambert. 1982. Virulent P. berghei malaria: prolonged survival and decreased cerebral pathology in cell-dependent nude mice. J. Immunol. 129:2213-2218.
16. Fritsche, G., C. Larcher, H. Schennach, and G. Weiss. 2001. Regulatory interactions between iron and nitric oxide metabolism for immune defense against Plasmodium falciparum infection. J. Infect. Dis. 183:1388-1394.
17. Hogg, N., V. M. Darley-Usmar, M. T. Wilson, and S. Moncada. 1992. Production of hydroxyl radicals from the simultaneous generation of superoxide and nitric oxide. Biochem. J. 281(Part 2):419-424.
18. Jaramillo, M., I. Plante, N. Ouellet, K. Vandal, P. A. Tessier, and M. Olivier. 2004. Hemozoin-inducible proinflammatory events in vivo: potential role in malaria infection. J. Immunol. 172:3101-3110.
19. Krishnegowda, G., A. M. Hajjar, J. Zhu, E. J. Douglass, S. Uematsu, S. Akira, A. S. Woods, and D. C. Gowda. 2005. Induction of proinflammatory responses in macrophages by the glycosylphosphatidylinositols (GPIs) of Plasmodium falciparum: cell signaling receptors, GPI structural requirement, and regulation of GPI activity. J. Biol. Chem. 280:8606-8616.
20. Krogstad, D. J. 1996. Malaria as a reemerging disease. Epidemiol. Rev. 18:77-89.
21. Liu, X., M. J. Miller, M. S. Joshi, H. Sadowska-Krowicka, D. A. Clark, and J. R. Lancaster, Jr. 1998. Diffusion-limited reaction of free nitric oxide with erythrocytes. J. Biol. Chem. 273:18709-18713.
22. Murphy, M. E., and E. Noack. 1994. Nitric oxide assay using hemoglobin method. Methods Enzymol. 233:240-250.
23. Nalwaya, N., and W. M. Deen. 2004. Analysis of the effects of nitric oxide and oxygen on nitric oxide production by macrophages. J. Theor. Biol. 226:409-419.
24. Nalwaya, N., and W. M. Deen. 2004. Peroxynitrite exposure of cells cocultured with macrophages. Ann. Biomed. Eng. 32:664-676.
25. Ockenhouse, C. F., S. Schulman, and H. L. Shear. 1984. Induction of crisis forms in the human malaria parasite Plasmodium falciparum by gamma-interferon-activated, monocyte-derived macrophages. J. Immunol. 133:1601-1608.
26. Postma, N. S., E. C. Mommers, W. M. Eling, and J. Zuidema. 1996. Oxidative stress in malaria; implications for prevention and therapy. Pharm. World Sci. 18:121-129.
27. Sanni, L. A., S. Fu, R. T. Dean, G. Bloomfield, R. Stocker, G. Chaudhri, M. C. Dinauer, and N. H. Hunt. 1999. Are reactive oxygen species involved in the pathogenesis of murine cerebral malaria J. Infect. Dis. 179:217-222.
28. Schofield, L., and F. Hackett. 1993. Signal transduction in host cells by a glycosylphosphatidylinositol toxin of malaria parasites. J. Exp. Med. 177:145-153.
29. Stevenson, M. M., and E. M. Riley. 2004. Innate immunity to malaria. Nat. Rev. Immunol. 4:169-180.
30. Taylor-Robinson, A. W., R. S. Phillips, A. Severn, S. Moncada, and F. Y. Liew. 1993. The role of TH1 and TH2 cells in a rodent malaria infection. Science 260:1931-1934.
31. van der Heyde, H. C., M. M. Elloso, J. vande Waa, K. Schell, and W. P. Weidanz. 1995. Use of hydroethidine and flow cytometry to assess the effects of leukocytes on the malarial parasite Plasmodium falciparum. Clin. Diagn. Lab. Immunol. 2:417-425.
32. van der Heyde, H. C., Y. Gu, Q. Zhang, G. Sun, and M. B. Grisham. 2000. Nitric oxide is neither necessary nor sufficient for resolution of Plasmodium chabaudi malaria in mice. J. Immunol. 165:3317-3323.
33. Vaughn, M. W., K. T. Huang, L. Kuo, and J. C. Liao. 2000. Erythrocytes possess an intrinsic barrier to nitric oxide consumption. J. Biol. Chem. 275:2342-2348.
34. Vaughn, M. W., L. Kuo, and J. C. Liao. 1998. Estimation of nitric oxide production and reaction rates in tissue by use of a mathematical model. Am. J. Physiol. 274:H2163-H2176.
35. Winterbourn, C. C. 1990. Oxidative reactions of hemoglobin. Methods Enzymol. 186:265-272.
36. Yoneto, T., T. Yoshimoto, C. R. Wang, Y. Takahama, M. Tsuji, S. Waki, and H. Nariuchi. 1999. Gamma interferon production is critical for protective immunity to infection with blood-stage Plasmodium berghei XAT but neither NO production nor NK cell activation is critical. Infect. Immun. 67:2349-2356.(Peter Sobolewski, Irene G)