- Email: [email protected]
Involvement of nitric oxide (NO) and TNF-α in the oxidative stress associated with anemia in experimental Trypanosoma cruzi infection
FEMS Immunology and Medical Microbiology 41 (2004) 69–77 www.fems-microbiology.org
Involvement of nitric oxide (NO) and TNF-a in the oxidative stress associated with anemia in experimental Trypanosoma cruzi infection Aparecida Donizette Malvezi a, Rubens Cecchini a, Fausto de Souza a, Carlos Eduardo Tadokoro b, Luiz Vicente Rizzo b,c,d, Phileno Pinge-Filho a
d
a,*
Department of Pathological Sciences, State University of Londrina, CEP:86051-970. Parana, Brazil b Department of Immunology, University of S~ao Paulo, S~ao Paulo, Brazil c F undacß~ao Zerbini, S~ao Paulo, Brazil University of S~ao Paulo Medical School, Div. of Allergy and Clinical Immunology, S~ao Paulo, Brazil Received 14 July 2003; received in revised form 18 November 2003; accepted 12 January 2004 First published online 5 February 2004
Abstract Trypanosoma cruzi infection in mice is associated with severe hematological changes, including anemia, which may contribute to mortality. TNF-a and nitric oxide (NO) play a critical role in establishing host resistance to this pathogen. We hypothesized that phagocyte-derived NO damages erythrocytes and contributes to the anemia observed during T. cruzi infection. To test this hypothesis, two strains of mice that differed in susceptibility and NO response to T. cruzi infection were used in these studies. We also blocked endogenous NO production by aminoguanidine (AG) treatment or blocked TNF-a with a neutralizing antibody and used mice that cannot produce phagocyte-derived NO (C57BL/6 iNOS= ). Following infection with T. cruzi, resistant (C57BL/6) and susceptible (Swiss) mice displayed a parasitemia that peaked at the same time (i.e., day 9), yet parasitemia was 3-fold higher in Swiss mice (P < 0:05). All Swiss mice were dead by day 23 post-infection, while no C57BL/6 mice died during the study. At 14 days postinfection anemia in C57BL/6 mice was more severe than in Swiss mice. Treatment of both strains with the NO inhibitor, AG (50 mg/ kg), and the use of iNOS= mice, revealed that the anemia in T. cruzi-infected mice is not caused by NO. However, the reticulocytosis that occurs during infection was significantly reduced after treatment with AG in both Swiss and C57BL/6 mice (P < 0:05). In addition, we showed that neutralization of TNF-a in vivo induced a significant increase in circulating reticulocytes in T. cruziinfected C57BL/6 mice (P < 0:05), but did not modify other hematologic parameters in these mice. The evaluation of the oxidative stress after induction by t-butyl hydroperoxide (t-BHT) revealed that the treatment with AG completely protected against NOmediated haemoglobin oxidation. Further, treatment with AG, but not with anti-TNF-a, protected against the infection-induced reduction of antioxidant capacity of erythrocytes as assessed by oxygen uptake and induction time. In summary, this is the first report showing the participation of NO and TNF-a in the oxidative stress to erythrocytes in acute T. cruzi infection. Further, our data suggest that NO does not play a direct role in development of the anemia. However, NO may contribute to other hematological changes noted during T. cruzi infection, such as the elevation of circulating reticulocytes and the reduction in circulating leukocytes and neutrophils. 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Trypanosoma cruzi; ChagasÕ disease; Anemia; Oxidative stress; Nitric oxide; TNF-a
1. Introduction *
Corresponding author. Tel.: +55-43-3371-4267; fax: +55-43-33714207. E-mail address: [email protected] (P. Pinge-Filho).
ChagasÕ disease or American trypanosomiasis is caused by the protozoan Trypanosoma (Schizotrypanum) cruzi [1]. The infection is endemic in Latin
0928-8244/$22.00 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsim.2004.01.005
70
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
America and serological data indicate that over 16 million people are infected [2]. In his pioneering work, Carlos Chagas was first to report anemia in patients suffering from a T. cruzi infection [1]. Since the discovery of ChagasÕ disease over 90 years ago, few studies have explored the causes of anemia associated with T. cruzi infection. Mice inoculated with different strains of the T. cruzi present intense thrombocytopenia [3] and neutropenia following by neutrophilia and eosinophilia [4]. Similar hematological alterations have been also described in experimental African trypanosomiasis [5] and is a common characteristic of human immunodeficiency virus infection [6] and malaria [7,8]. Recently, Marcondes and collaborators [9] showed that experimental acute T. cruzi infection is associated with anemia, thrombocytopenia, leukopenia, and bone marrow hypoplasia and that these alterations can be prevented by nifurtimox (an anti-trypanosomal drug) treatment. However, the mechanisms responsible for these hematological alterations are not well understood. Our hypothesis is that cytokines and nitric oxide secreted during the acute phase of a T. cruzi infection [10] are responsible for the hematological changes observed. IFN-c and TNF-a are among the cytokines that play an important role in host defense against T. cruzi [11]. TNF-a can modulate myelopoiesis, erythropoiesis, lymphopoiesis and thrombocytopoiesis [12]. In malaria, TNF-a has been shown to be an important mediator of anemia [13]. However, no such relationship has been established for TNF-a and the anemia associated with T. cruzi infection. Host resistance to T. cruzi is dependent, in part, upon the expression of inducible NO synthase (iNOS) [14] in macrophages. IFN-c and TNF-a play a central role in stimulating iNOS expression and subsequent nitric oxide ( NO) production by macrophages [14]. However, reactive nitrogen species, such as NO, along with reactive oxygen species (ROS) produced by activated macrophages have been implicated as mediators of tissue damage in numerous pathological conditions [15,16]. Erythrocytes are among the tissues that NO can injury [17,18]. For instance, it has been suggested that NO-mediated deformability of erythrocytes plays a role in sepsis and systemic inflammatory response syndrome [19]. Furthermore, NO can cross the erythrocyte membrane and react with SH groups mainly with haemoglobin-cysteine groups [20]. This reaction yields two products: S-nitroso-haemoglobin (SNOHb) and nitrosyl-haemoglobin (HbNO) [21,22]. The nitrosyl complex slowly oxidizes to methaemoglobin (metHb) in presence of oxygen [22], thus measuring metHb can be a useful index of in vivo NO-mediated oxidative stress [23]. NO produced by macrophages reacts very fast with super oxide anion (O 2 ) yielding peroxynitrite (ONOO ), which is a powerful oxidizing and lipid peroxidizing agent [15].
In murine leishmaniasis, there is evidence of enhanced levels of NO in the blood of infected mice and of peroxynitrite associated with their lesions [24]. The induced damage depend on the balance between NO and O 2 . For example, equal fluxes of O 2 and NO can stimulate lipid peroxidation, a high ratio of NO to O 2 can decrease the peroxidation because of the ability of NO to scavenge peroxyl and alkoxyl radicals [15]. Therefore, we predict that the nature and extent of erythrocyte oxidative injury during infection with T. cruzi will depend upon three factors: (1) the induction of iNOS and thus production of NO in response to the infection; (2) the oxidative stress generated outside of erythrocyte, particularly phagocyte-derived O 2 ; (3) the rate of reaction between NO and either haemoglobin or O 2 . Since in T. cruzi infection the concentration of both NO and O 2 are enhanced [25,26], we hypothesized that endogenous NO is responsible for the hematological changes (i.e., anemia) observed in the acute phase of T. cruzi infection. Therefore, in order to test the possible involvement of NO as a part of injury mechanism in experimental ChagasÕ disease, we carried out studies with two strains of mice that differed in iNOS induction following T. cruzi infection. C57BL/6 mice are resistant to T. cruzi and respond to infection with a strong induction of the iNOS gene in phagocytes, while Swiss mice are susceptible to this parasite and have a more modest induction of iNOS following infection. The importance of endogenous NO production on the hematological changes associated with T. cruzi infection was also tested by using C57BL/6iNOS= mice deficient in iNOS and by treating wild-type C57BL/6 mice with aminoguanidine (AG), an inhibitor of iNOS. We measured haemoglobin oxidation and assessed oxidative stress by measuring tbutyl hydroperoxide (t-BHP) induced oxygen uptake and induction time of isolated erythrocytes. The possible involvement of TNF-a in the anemia associated with T. cruzi infection was investigated by treating wild-type C57BL/6 mice with a neutralizing anti-TNF-a antibody. Our results clearly show that NO produced during T. cruzi infection increases erythrocyte oxidative stress, but is not necessary for the development of anemia in acute T. cruzi infection in mice.
2. Materials and methods This study was reviewed and approved by the Internal Scientific Commission and the Bioethics in Research Committee of the Londrina State University, Londrina, Brazil. 2.1. Mice and Trypanosoma cruzi infection Swiss mice were obtained from the breeding colonies of the animal facility of the Center for Biological Sciences
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
at Londrina State University, Londrina, Brazil. C57BL/6 mice and C57BL/6 iNOS= (mice deficient in iNOS) [27], were obtained from the mouse breeding facilities of the Department of Immunology, Institute of Biomedical Sciences, University of S~ ao Paulo, S~ ao Paulo, Brazil. We used 8 to 12 weeks old male mice in all experiments. Mice were maintained under standard conditions in an animal facility within the Department of Pathological Sciences, Center for Biological Sciences, Londrina State University. Commercial rodent diet (Nuvilab-CR1, Nuvital, Campo Mour~ ao, Brazil) and sterilized water were available ad libitum. All procedures with the animals were in accordance with the guidelines of the Brazilian Code for the Use of Laboratory Animals. The Y strain of T. cruzi was kindly supplied by Dr. Paulo Maria Ferreira de Ara ujo (Institute of Biosciences, Campinas State University, Campinas, S~ao Paulo, Brazil). This strain was maintained by weekly intraperitoneally (i.p.) inoculation of Swiss mice with 2 105 blood trypomastigotes forms. Infective blood trypomastigotes were obtained from T. cruzi-infected mice by drawing blood via cardiac puncture following anaesthetization. Motile blood forms were counted and the desired number of parasites (5 · 103 ) was injected intraperitoneally (i.p.) in Swiss, C57BL/6 and C57BL/6 iNOS= mice. Parasitemia was assessed by counting circulating parasites in 5 ll of blood obtained from a tail vein of infected mice. These data were expressed as the number of parasites per milliliter of blood [28]. Parasitemia and survival rates were determined daily, beginning at the 5th day of infection. 2.2. Treatment of infected mice with aminoguanidine During the first 15 days of infection, Swiss and C57BL/6 mice received daily i.p. injections of saline with or without aminoguanidine (Sigma, St. Louis), an inhibitor of inducible nitric oxide synthase (iNOS), diluted in sterile phosphate-buffered saline (50 mg/kg of body weight). The first dose was given 4 h after T. cruzi infection. The dose of AG chosen for these experiments was based on previously published studies demonstrating its efficacy [29].
71
cardiac puncture with heparinized needles and syringes and counted by standard methods [30]. Haemoglobin concentration was determined by the Drabkin method. Hematocrits were obtained by microcentrifugation of capillary tubes filled with heparinized blood [31]. For enumerating reticulocytes, 40 ll of blood heparinized was incubated with 20 ll brilliant cresyl blue for 20 minutes at 37 C, then thin blood smears were prepared on glass slides. After the blood smears air dried, reticulocytes were counted by light microscopy. All blood analysis and cell counts were performed 0, 7 and 14 days post-infection.
2.5. Bone marrow cell harvest Bone marrow cells were harvested by flushing the femoral shafts with ice-cold PBS, as previously described [32]. The total number of cells collected was determined by manual hemacytometer count. For differential counts, cell suspensions from uninfected and day 14-infected mice were deposited on glass slide, stained with May Gr€ unwald-Giemsa, then counted by light microscopy.
2.6. Determination of oxygen uptake and induction time (T ind ) in erythrocytes Heparinized blood samples from uninfected and infected mice (day 14 post-infection) were used for erythrocyte oxidative stress determinations [15,33,34]. After removal of plasma and white cells from whole blood, the remaining erythrocytes were washed twice with a 10 mM sodium phosphate buffered saline (0.9% NaCl, pH 7.4), then resuspended in the same buffer (1:99, v/v). Both t-butyl hydroperoxide 2 mM (t-BHT) induced oxygen uptake and induction time were measured with a Clark-type oxygen electrode at 37 C [33,34]. The induction time is directly related to the intracellular protective antioxidant capacity, while oxygen uptake is an indirect measure of the susceptibility of erythrocytes membranes to lipid peroxidation elicited by t-BHT [15,34].
2.3. In vivo neutralization of TNF-a C57BL/6 mice were injected i.p with 200 lg of an anti-TNF-a mAb (XT22) [10], kindly provided by Dr. Ises A. Abrahamsohn, Department of Immunology, University of S~ ao Paulo, S~ ao Paulo, Brazil. XT22 was given on days 7, 9, 11 and 13 of T. cruzi infection. 2.4. Hematological methods Erythrocytes and leukocytes from normal and infected mice under ether anesthesia, were collected from
2.6.1. Haemoglobin oxidation measurement Oxidation of haemoglobin was determined using a dual beam spectrophotometer (Varian 634S). Erythrocytes were lysed with distilled-deionized water and the spectrum of a 1% solution of lysates was measured from 500 to 700 nm. The signal was digitized and analyzed with the Lynx program (Lynx Tecnologia Eletr^ onica, Ltda). The concentration of oxyhaemoglobin, metahaemoglobin and haemichrome were calculated using the Winterbourn equations [35].
72
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
2.7. Statistical analysis Data are expressed as mean standard deviation of the means. The impact of infection and other treatments were determined by two-way Analysis of Variance (ANOVA). When significant main effects were noted, differences between individual groups were tested using Bonferonni multiple comparisons method. Survival curves were compared using the Mantel-Haenszel log rank test. All statistical analysis was made using GraphPad Prism version 3.0 (GraphPad Software, San Diego, CA).
3. Results 3.1. Course of infection Five and seven days following inoculation with T. cruzi, parasitemia was similar between susceptible (Swiss) and resistant (C57BL/6) strains of mice (Fig. 1(a)). Parasitic load in the blood peaked 9 days post-infection and was significantly higher in Swiss compared to C57BL/6 mice. C57BL/6 iNOS= mice showed parasitemia levels more similar to the susceptible Swiss mice than their parental C57BL/6 strain. There were large and statistically significant differences (P < 0:001) in survival between all three strains of mice studied (Fig. 1(b)). All of the C57BL/6 mice survived the infection, while all of the Swiss and C57BL/6 iNOS= mice died by 23 and 15 days post-infection, respectively. 3.2. Development of anemia We conducted blood cells counts and measured several haematological values in uninfected and T. cruziinfected mice at 7 and 14 days post-infection. Based on significant decreases in haemoglobin, hematocrit and erythrocyte numbers, all three mouse strains studied suffered from anemia 14 days post-infection (Table 1). The severity of the anemia was greatest for C57BL/6 mice. Inhibition of endogenous NO production via AG treatment and neutralizing TNF-a did not impact the extent of anemia that developed 14 days post-infection (Table 1). Both 7 and 14 days post-infection we noted a significant increase in reticulocytes (P < 0:001) present in both the Swiss and C57BL/6 mice (Fig. 2(a) and (b)). Treatment of Swiss mice with AG completely abrogated this increase in reticulocytes associated with T. cruziinfection. Similarly, AG treatment of C57BL/6 mice significantly reduced the reticulocytosis associated with T. cruzi-infection in this resistant strain of mice. In the C57BL/6 iNOS= mice we noted a transient reticulocytosis, with a significant elevation in reticulocytes at 7, but not at 14 days post-infection (Fig. 2(b)). Surprisingly, we observed a very large increase in reticulocytes
Fig. 1. Parasitemia (a) and survival (b) of mice infected with T. cruzi. Swiss, C57BL/6 and C57BL/6 iNOS= mice were infected with 5 103 blood trypomastigotes Y strain T. cruzi. Parasitemia was quantified as trypomastigotes per milliliter of blood. Results are expressed as the mean standard deviation from of 10 mice per group, in an experiment representative of three similar experiments. Results were analysed by analysis of variance (ANOVA) followed by DunnÕs nonparametric test. * P < 0:05, significant difference in parasitemia, C57BL/6 and C57BL/6 KO iNOS= vs Swiss. ** Survival curve significantly different from the of C57BL/6 mice, P < 0:01.
in the T. cruzi-infected mice treated with the neutralizing antibody against TNF-a (Fig. 2(b); P < 0:001). 3.3. Changes in blood leukocytes and neutrophils Fourteen days following T. cruzi-infection, Swiss mice presented with neutropenia (50% decline; P < 0:05) without changes in mononuclear leukocyte numbers, while C57BL/6 mice showed a 50% decline in leukocytes (P < 0:05) and no change in neutrophils (P > 0:05) (Table 1). Treatment of C57BL/6 mice with AG tended to diminish the decline in leukocyte numbers associated with infection. In contrast, AG treatment in Swiss mice significantly reduced leukocyte numbers (25–30%; P < 0:05) in uninfected and infected mice alike. Treatment of Swiss mice also modestly reduced blood neutrophils in uninfected mice (P < 0:05), but it was able to partially restore neutrophil counts following T. cruzi-infection (P < 0:05). Neutralization of endoge-
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
73
Table 1 Hematological values on day 14 in T. cruzi-infected micea Group
Haemoglobin (g/dl)
Hematocrit (%)
Erythrocytes (109 )
Leukocytes (106 )
Neutrophils (106 )
Swiss C C + AG I I + AG
13.8 0.5 13.7 1.3 10.8 0.9 10.3 0.6
44.9 1.3 42.6 0.5 35.5 3.8 32.4 2.0
7.0 0.5 7.9 0.2 6.4 0.5 6.1 0.5
5.4 0.7 4.0 0.6 5.4 1.5 3.7 0.5
1.2 0.3 0.9 0.2 0.5 0.3 0.9 0.2
C57BL/6 C C + AG I I + AG I + Anti-TNF
13.5 0.7 12.6 0.9 8.2 0.7 8.6 0.9 8.7 0.7
44.0 0.8 40.3 3.5 28.1 1.4 28.3 2.6 28.3 2.3
7.6 0.8 7.0 0.3 5.1 0.5 5.3 0.3 5.0 0.3
5.0 0.7 5.4 1.2 3.0 0.8 4.0 0.7 5.0 0.8
0.4 0.02 0.5 0.1 0.4 0.1 1.1 0.1 1.3 02
INOS-/C I
12.7 1.1 9.3 1.2
40.5 3.4 29.8 2.4
7.4 0.5 5.2 0.6
5.4 1.2 2.6 0.7
1.6 0.2 0.4 0.1
a Groups of mice were infected (I) with 5 103 T. cruzi or infected and treated with aminoguanidine (AG) or anti-TNF-a mAb. (c) control; (C + AG) Control and AG-treated; (I) infected; (I + AG) infected and treated with AG. Values represent the mean SD and are representative of two independent experiments, using 5–12 mice per group. * Significantly different (P < 0:05) when compared with non-infected group. ** Significantly different (P < 0:001) when compared with non-infected group.
marked leukopenia and neutropenia (i.e., a 50% and 75% decline, respectively; P < 0:05). 3.4. Changes in cellularity and erythroblasts in the bone marrow
Fig. 2. Percentage of reticulocytes in the blood. Swiss mice (a), C57BL/ 6 and iNOS= mice (b) were infected with 5 103 T. cruzi. Four hours following infection and then daily for 15 days some of the mice in each strain received an i.p. injection of aminoguanidine (AG, 50 mg/kg BW) or 200 lg of anti-TNFa mAB on days 7, 9, 11 and 13 post-challenge. Control mice received saline injections only. Values represent the mean standard deviation and are representative of two independent experiments, with 5–12 mice per group. Differences in percentage of reticulocytes were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, P < 0:05.
nous TNF-a was able to completely inhibit the decline in blood leukocytes associated with T. cruzi-infection in C57BL/6 mice. The C57BL/6 iNOS= mice showed a
T. cruzi infection in Swiss and C57BL/6 mice was associated with a significant decrease in total number of bone marrow cells, in particular erythroblasts (Fig. 3(a)– (d); P < 0:05). AG treatment had no significant effect on the hypoplasia in the bone marrow of Swiss mice (Fig. 3(a) and (c)). In contrast, AG treatment of C57BL/ 6 mice prevented some of the decline in bone marrow cellularity associated with T. cruzi infection (Fig. 3(b) and (d); P < 0:001). However, the T. cruzi infectioninduced decline in total cells and erythroblasts in the bone marrow of C57BL/6 iNOS= was nearly identical to that observed in C57BL/6 mice (Fig. 3(b) and (d)). Neutralization of endogenous TNF-a in C57BL/6 mice prevented the decline in erythoblasts without significantly changing the decline in total cellularity in the bone marrow of infected mice (Fig. 3(d)). 3.5. Erythrocyte oxidative stress The ratio of oxyhaemoglobin to methaemoglobin (i.e., % oxidation) was increased following T. cruzi infection in both Swiss and C57BL/6 mice (Table 2). The extent of oxidation tended to be higher in Swiss compared to C57BL/6 mice. As expected, treatment with AG completely protected against this NO-mediated oxidation of haemoglobin. We noted an increase in the % oxidation of haemoglobin in C57BL/6 iNOS= mice, this appeared to be primarily a consequence of the
74
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
Fig. 3. Bone marrow cells and erythroblasts in T. cruzi-infected mice. Swiss mice (a), C57BL/6 and iNOS= mice (b) were infected with 5 103 T. cruzi. Four hours following infection and then daily for 15 days some of the mice in each strain received an i.p. injection of aminoguanidine (AG, 50 mg/kg BW) or 200 lg of anti-TNFa mAB on days 7, 9, 11 and 13 post-challenge. Control mice received saline injections only. Values represent the mean standard deviation and are representative of two independent experiments, with 5–12 mice per group. Differences in percentage of reticulocytes were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, P < 0:05. Table 2 Oxyhaemoglobin (OxyHb) oxidation ratea in T. cruzi-infected miceb Group
OxyHb (lM)
MetaHb + Hemic. (lM)
Rate oxidation (%)
Swiss C C + AG I I + AG
19.5 22.0 17.7 19.4
5.6 6.8 8.7 4.5
28.7 30.1 49.1 23.1
C57BL/6 C C + AG I I + AG I + Anti-TNF
24.0 20.8 16.3 18.8 14.8
5.2 3.0 5.2 4.6 4.8
21.6 14.4 31.9 24.4 32.4
INOS-/C I
29.2 17.9
8.2 6.7
28.0 37.4
a Oxyhaemoglobin (OxyHb) oxidation rate on day 14 of T. cruzi infection. b Groups of mice were infected (I) with 5 103 T. cruzi or infected and treated with aminoguanidine (AG) or anti-TNF-a mAb. (c) control; (C + AG) Control and AG treated; (I) infected; (I + AG) infected and treated with AG. Values are representative of two independent experiments, using 5–12 mice per group.
infection-induced anemia with the associated decline in haemoglobin levels. Anti-TNF-a treatment did not affect haemoglobin oxidation in C57BL/6 mice infected with T. cruzi. Oxygen uptake by erythrocytes is directly associated with the susceptibility of the erythrocyte membrane to undergo lipid peroxidation elicited by t-BHP and is
proportional to previous oxidative stress experienced by the erythrocyte in vivo. The oxygen uptake by erythrocytes from Swiss mice was significantly increased 14 days following infection (Fig. 4(a); P < 0:05). AG treatment significantly diminished erythrocyte oxygen uptake in both uninfected and infected Swiss mice (Fig. 4(a); P < 0:001). In C57BL/6 mice T. cruzi infection was associated with a significant decrease in oxygen uptake at 14 days post-infection (Fig. 4(b); P < 0:001). AG treatment enhanced oxygen uptake in uninfected mice, but diminished it following infection. Prior to infection, iNOS= mice showed very low levels of oxygen uptake, which increased dramatically following infection. Treatment with anti-TNF-a did not alter the oxygen uptake profile of erythrocytes from C57BL/6 mice prior to or after T. cruzi infection. Induction time (Tind ) is directly related to the intracellular protective antioxidant capacity of the erythrocyte. Neither infection nor AG treatment altered Tind in Swiss mice (P > 0:05, data not shown). In contrast, T. cruzi infection in the C57BL/6 mice resulted in a significant reduction in Tind (Fig. 5; P < 0:05). Treatment with AG completely abrogated the infection associated decline in Tind (P < 0:01, Fig. 5). Prior to infection, Tind was significantly longer in erythrocytes isolated from C57BL/6 iNOS= mice compared to C57BL/6 mice. However, like wild-type C57BL/6 mice, a significant decrease in Tind was seen for the C57BL/6 iNOS= 14 days post-infection (P < 0:001). Treatment with antiTNF-a also protected against the reduction of antioxidant capacity of C57BL/6 erythrocytes on day 14 post-infection (Fig. 5; P < 0:001).
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
Fig. 4. Erythrocyte oxidative stress during T. cruzi infection. Here we measured oxygen uptake induced by t-BHT in erythrocyte from Swiss mice (a) and from C57BL/6 and iNOS= mice (b). Four hours following infection and then daily for 15 days some of the mice from strain received an i.p. injection of aminoguanidine (AG, 50 mg/kg BW) or 200 lg of anti-TNFa mAb on day 7, 9, 11 and 13 post-challenge. Controls received saline injections only. Values represent the mean standard deviation and are representative of two independent experiments, using 5 mice per group. Differences in oxygen uptake were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, *P < 0:05.
Fig. 5. Erythrocyte antioxidant capacity following T. cruzi infection. We measured induction time (Tind ) following t-BHT treatment of erythrocytes from C57BL/6 and C57BL/6 iNOS= mice. Mice were treated as described in the previous figure. Values represent the mean standard deviation and are representative of two independent experiments, using 5 mice per group. Differences in induction time levels were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, *P < 0:05.
4. Discussion Infection of mice with T. cruzi is characterized by exponential growth of parasites and anemia in both resistant and susceptible strains of mice. In this study we demonstrated that T. cruzi infection is lethal in susceptible mouse strains (i.e., Swiss and C57BL/6 iNOS=
75
mice). Similar results have been reported previously [3,36]. Resistance to T. cruzi infection in mice depends on the production of TNF-a and NO [37]. However, excessive NO production may result in severe tissue damage, including erythrocyte destruction. Acute infection with T. cruzi has been associated with significant alterations in hematological values in the blood and bone marrow [9]. We showed that compared to susceptible mice, T. cruzi infection in resistant mice resulted in a more severe anemia (i.e., reduced Hb and Hct). Infection was also associated with a time-dependent increase in blood reticulocytes and a decrease in bone marrow erythroblasts (two markers of anemia). Consistent with our Hb and Hct data, the changes in reticulocytes and erythroblasts tended to be greater in resistant (C57BL/6) versus susceptible mice (Swiss). The primary objective of this study was to determine if endogenous NO production was responsible for the hematological changes (i.e., anemia) observed in a T. cruzi infection. Blocking endogenous NO production via AG treatment had no discernable impact on the infection-induced changes in hematological parameters measured in this study for either susceptible or resistant mouse strains. We did not directly measure in vivo NO production, however, the ability of AG treatment to abrogate NO-mediated oxidation of haemoglobin during the infection supports the conclusion that AG was effectively inhibiting NO production in vivo. The occurrence of anemia in C57BL/6 iNOS= mice following infection with T. cruzi provides further evidence that NO is not required for the development of infectioninduced anemia. However, our data clearly demonstrate that endogenous NO production is essential for the control of parasitemia and survival following a T. cruzi infection. Similar results have been reported previously [36]. AG treatment was able to modify two indirect indicators of anemia. First, AG treatment prevented the T. cruzi infection-induced increase in blood reticulocytes in both susceptible and resistant strains of mice. Second, the decrease in bone marrow erythroblasts that occurred 14 days post-infection was not as large if mice were treated with AG. Also, AG treatment was able to prevent neutropenia in Swiss mice and diminish the leukopenia in C57BL/6 mice at day 14 post-infection, probably due a reduction in the cellular toxicity provoked by NO. In addition, we showed that in vivo neutralization of TNF-a with monoclonal antibody (XT22) treatment, provoked a dramatic increase in the percentage of reticulocytes in C57BL/6 mice 14 days after T. cruzi infection, but didnÕt modify the anemia of those animals. The leukopenia observed in infected C57BL/6 mice was abolished for the treatment with anti-TNF-a. These data suggest that the TNF-a produced by activated macrophages during the acute infection with the T. cruzi
76
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
diminishes erythropoiesis. In fact, previous studies have demonstrated inhibitory effects of TNF-a on erythropoiesis [12,38]. In this study we assessed oxidative stress in erythrocytes three ways: by measuring oxyhaemoglobin and haemoglobin derivatives (i.e., methaemoglobin and hemichrome), oxygen uptake and induction time (Tind ) following treatment with t-butyl hydroperoxide, t-BHT. We found that an acute infection of mice with T. cruzi was generally associated with increased erythrocyte oxidative stress. For example, we observed a modest, but significant increase the ratio of metHb and hemichrome over oxyHb following T. cruzi infection in Swiss and C57BL/6 mice. Further, we found evidence of lipid peroxidation in erythrocytes (i.e., increased oxygen uptake) from T. cruzi-infected Swiss mice. That AG treatment reduced erythrocyte oxygen uptake both prior to and following infection suggested that NO played a role in lipid peroxidation of erythrocyte membranes. Surprisingly, and in contrast to our observations in Swiss mice, T. cruzi infection in C57BL/6 mice significantly reduced in vitro oxygen uptake by erythrocytes. However, blocking in endogenous NO production in C57BL/6 mice by AG treatment significantly reduced oxygen uptake by erythrocytes just as it had done in Swiss mice. We do not know for certain how infection causes oxidative stress in erythrocytes, but we believe that phagocyte-derived NO and O are central to this 2 process. O can react with NO yielding ONOO or by 2 the Fenton reaction produce OH, both of which could lead to oxidative stress in erythrocytes. Since Swiss mice are only modestly iNOS responsive, in vivo NO formation following T. cruzi infection is expected to be lower is Swiss than C57BL/6 mice. With only modest NO production, we believe that a larger fraction of NO in Swiss mice will be present as HbFe(II)(NOþ ) and consequently lead to higher relative yields of nitrosylhaemoglobin (SNOHb) [22]. The nitrosyl-haemoglobin complex slowly oxidizes to metHb, which is known to reduce the flux of O 2 [22]. The net result should be lower internal oxidative stress in erythrocytes from Swiss than C57BL/6 mice. Our observation that antioxidant capacity as evaluated by Tind was not significantly affected by T. cruzi infection in Swiss mice is consistent with this theoretical mechanism. Our results obtained with C57BL/6 mice suggested that the nature of the oxidative stress that occurred during an acute T. cruzi infection was somewhat different than that found in erythrocytes from Swiss mice. In particular, the greater iNOS induction seemed to contribute to a higher oxidation rate of Hb than we observed in Swiss mice. However, similar to Swiss mice, prior in vivo AG treatment of C57BL/6 mice abrogated most of this infection-induced Hb oxidation and diminished erythrocyte oxygen uptake as well. Unlike in
Swiss mice, our data in C57BL/6 mice suggest that oxidation of Hb to metHb does diminish the antioxidant defense capacity of erythrocytes. It has been shown that besides intracellular GSH, HbSH groups contribute to the antioxidant capacity of erythrocytes [23]. These data suggests that when NO is abundant intracellular oxidative stress is favored over oxidative attack of the erythrocyte membrane. In fact in the complete absence of NO, membrane injury as assessed by oxygen uptake was greatly increased in erythrocytes isolated from C57BL/6 iNOS= 14 days after infection. These results suggest that participation of NO in membrane oxidative injury may be indirect. Altogether our results indicate that in experimental ChagasÕ disease oxidative stress inside and outside of the erythrocyte occurs, but that the precise nature of this oxidative stress may depend on the balance of NO/O 2 produced by the host. We conclude that death associated with an acute T. cruzi infection is not caused by NO-induced anemia or alterations in bone marrow cellularity. Our data confirm the importance of NO as part of the hostÕs defense against T. cruzi infection. However, the impact of NO on cells of the blood and bone marrow appears to be complex, and NO-based therapies against infections are not easy to design and demand caution.
Acknowledgements We thank Vargas, J.A. for excellent technical assistance, Araujo, P.M.F (State University of Campinas) for providing T. cruzi (Y strain). We are also grateful to Abrahamsohn, I.A for anti-TNF-a mAb (University of S~ao Paulo) and Dr. Kevin L. Fritsche (University of Missouri) for his assistance with editing the manuscript. Fausto de Souza was the recipient of a CNPq/PIBIC/ UEL undergraduate fellowship.
References [1] WHO – World Health Organization (1997). Tropical disease research: progress report 95–96. p. 134. [2] Chagas, C. (1909) Nova tripanozomiase humana. Estudos sobre a morfologia e o ciclo evolutivo do Schizotrypanum cruzi n.gen., s.sp. agente etiol ogico de uma nova entidade m orbida do homem. Mem. Inst. Oswaldo Cruz 1, 159–218. [3] Cardoso, J.E. and Brener, Z. (1980) Hematological changes in mice experimentally infected with Trypanosoma cruzi Mem. Inst. Oswaldo Cruz 75, 97–104. [4] Repka, D., Rangel, H.A., Atta, A.M., Gavino, V.A. and Piedrabuena, A.E. (1985) Experimental Chagasas diseases in mice infected with one LD50 of parasite. Rev. Bras. Biol. 45, 309–316. [5] Ikede, B.O., Lule, M. and Terry, R.J. (1977) Anemia in trypanosomiasis: mechanisms of erythrocyte destruction in mice infected with Trypanosoma congolense or T. bruce i. Acta Trop. 34, 53–60.
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77 [6] Claster, S. (2002) Biology of anemia, differential diagnosis, and treatment options in human immunodeficiency virus infection. J. Infect. Dis. 85, 105–109. [7] Weatherall, D.J., Miller, L.H., Baruch, D.I., Marsh, K., Doumbo, O.K., Casals-Pascual, C. and Roberts, D.J. (2002) Malaria and red cell. Hematology (Am. Soc. Hematol. Educ. Program), 35–57. [8] Paul, R.E. and Brey, P.T. (2003) Malaria Parasites and red blood cells: from anaemia to transmission. Mol. Cells 15, 139–149. [9] Marcondes, M.C., Borelli, P., Yoshida, N. and Russo, M. (2000) Acute Trypanosoma cruzi infection is associated with anemia, thrombocytopenia, leukopenia, and bone marrow hypoplasia: reversal by nifurtimox treatment. Microbes Infect. 4, 347–352. [10] Abrahamsohn, I.A. and Coffman, R.L. (1995) Cytokine and nitric oxide regulation of the immunosuppression in Trypanosoma cruzi infection. J. Immunol. 155, 3955–3963. [11] Munoz-Fernandez, M.A., Fernandez, M.A. and Fresno, M. (1992) Synergism between tumor necrosis factor-alpha and interferon-gamma on macrophage activation for the killing of intracellular Trypanosoma cruzi through a nitric oxide-dependent mechanism. Eur. J. Immunol. 22, 301–307. [12] Ulich, T.R., Shin, S.S. and del Castillo, J. (1993) Haematologic effects of TNF. J. Res. Immunol. 144, 347–354. [13] Miller, K.L., Silverman, P.H., Kullgren, B. and Mahlmann, L.J. (1989) Tumor necrosis factor alpha and the anemia associated with murine malaria. Inf. Immun. 5, 1542–1546. [14] MacMicking, J., Xie, Q.W. and Nathan, C. (1997) Nitric oxide and macrophage function. Annu. Rev. Immunol. 5, 323–350. [15] Halliwell, B. and Gutteridge, J.M. (1999) Free Radicals, Other Reactive Species and Disease. In: Free radicals in biology and medicine, pp. 660–699. Oxford University Press. [16] Oliveira, F.J.D.A. and Cecchini, R. (2000) Oxidative stress of liver in hamsters infected with Leishmania (L) chagasi. J. Parasitol. 86, 1067–1072. [17] Biswas, T., Pal, J.K., Naskar, K., Ghosh, D.K. and Ghosal, J. (1995) Lipid peroxidation of erythrocytes during anemia of the hamsters infected with Leishmania donovani. Mol. Cel. Bioch. 146, 99–105. [18] Biswas, T., Ghosh, D.K., Mukherjee, N. and Ghosal, J. (1997) Lipid peroxidation of erythrocytes in visceral leishmaniasis. J. Parasitol. 83, 151–152. [19] Baterman, R.M., Jagger, J.E., Sharpe, M.D., Ellsworth, M.L., Mehta, S. and Ellis, C.G. (2001) Erythrocyte deformability is a nitric oxide-mediated factor in decreased capillary density during sepsis. Am. J. Physiol. Heart Circ. Physiol. 280, 2848–2856. [20] Baskurt, O.K., Temiz, A. and Meiselman, H.J. (1998) Effect of superoxide anions on red blood cell rheologic properties. Free Radic. Biol. Med. 24, 102–110. [21] Rassaf, T., Bryan, N.S., Maloney, R.E., Speciean, V., Kelm, M., Kalyanaraman, B., Rodrigues, J. and Feelisch, M. (2003) NO adducts in mammalian red blood cells: too much or too little. Nat. Med. 9, 481–483. [22] Herold, S. and Rock, G. (2003) Reactions of Deoxy-Oxy-, and Methaemoglobin with Nitrogen Monoxide. J. Biol. Chem. 278, 6623–6634. [23] Rossi, R., Barra, D.B.A., Boumis, G., Canofeni, S., Simplicio, P.D., Lusini, L., Pascarella, S. and Amiconi, G. (1998) Fastreacting thiols in rat haemoglobins can intercept damaging species in erythrocytes more efficiently than glutathione. J. Biological Chemistry 273, 19198–19206.
77
[24] Augusto, O., Linares, E. and Giorgio, S. (1996) Possible roles nitric oxide and peroxynitrite in murine leishmaniasis. Bras. J. Med. Biol. Res. 29, 853–862. [25] Russo, M., Starombinas, N., Ribeiro-Dos-Santos, R., Minoprio, P. and Hontebeyrie-Joskowicz, M. (1989) Suscetible mice present higher macrophages activation than resistant mice during infections with myotropc strains of Trypanosoma cruzi. Parasite Immunol. 11, 385–395. [26] Cardoni, R.L., Rottemberg, M.E. and Segura, E.L. (1990) Increased production of reactive oxygen species by cells from mice acutely infected with Trypanosoma cruzi. Cell. Immunol. 1, 11–21. [27] Macmicking, J.D., Nathan, C., Hom, G., Chartrain, N., Fletcher, D.S., Trumbauer, M., Stevens, K., Xie, Q., Sokol, K., Hutchinson, N., Chen, H. and Mudgett, J.S. (1995) Altered responses to bacterial infection and endotoxic shock in mice lacking inducible Nitric oxide synthase. Cell 81, 641– 650. [28] Brener, Z. (1962) Therapeutic activity and criterion from cure on mice experimentally infected with Trypanosoma cruzi. Rev. Inst. Med. Trop. S~ao Paulo 4, 389–396. [29] Martins, G.A., Vieira, L.Q., Cunha, F.Q. and Silva, J.S. (1999) Gamma interferon modulates CD95 Fas and CD95 ligand (Fas-L) expression and nitric oxide-induced apoptosis during the acute phase of Trypanosoma cruzi infection: a possible role in immune response control. Infect. Immun. 67, 3864–3871. [30] Dacie, J.V. and Lewis, S.M. (1994) Basic haematological techniques. In: Practical Haematology (Dacie, J.V. and Lewis, S.M., Eds.), eighth ed. Churchill Livingstone, Edinburg. [31] Bain, B.J. (1997). Celulas Sanguıneas: Um guia pratico, 2nd ed., pp 25-51. Porto Alegre, Artes Medicas Sul LTDA. [32] Tadokoro, C.E. and Abrahamsohn, I.A. (2001) Bone marrowderived macrophages grown in GM-CSF or M-CSF differ in their ability to produce IL-12 and to induce IFN-gamma production after stimulation with Trypanosoma cruzi antigens. Immunol. Lett. 77, 31–38. [33] Lissi, E.R., Fransj, C., Fernandes, V. and Videla, L.A. (1986) Effects of antioxidants and haemoglobin status on the t-butyl hydeoperoxide-induced oxygen uptake by red blood cells. Cell. Biochem. Funct. 4, 447–454. [34] Bainy, A.C.D., Saito, E., Carvalho, P.S.M. and Junqueira, V.B.C. (1996) Oxidative stress in gill, erythrocytes, liver and kidney of Nile tilapia (Oreochromis niloticus) from a polluted site. Aquatic Toxicol. 34, 151–162. [35] Winterbourn, C.C. (1985). Reaction of superoxide with haemoglobin. CRC handbook of methods for oxigen radical research, pp 137–141. [36] Saeftel, M., Fleischer, B. and Hoerauf, A. (2001) Stage-dependent role of nitric oxide in control of Trypanosoma cruzi infection. Infect. Immun. 69, 2252–2259. [37] Silva, S.S., Vespa, G.N.R., Cardoso, M.A.G., Aliberti, J.C.S. and Cunha, F.Q. (1995) Tumor necrosis factor alfa mediates resistance to Trypanosom cruzi infection in mice by inducing nitric oxide production in infected gamma interferon-activated macrophages. Inf. Immun. 12, 4862–4867. [38] Ulich, T.R., Castillo, J. and Yin, S. (1990) Tumor necrosis factor exerts dose-dependent effects on erythropoiesis and myelopoiesis in vivo. Exp. Hematol. 18, 311–315.
Involvement of nitric oxide (NO) and TNF-a in the oxidative stress associated with anemia in experimental Trypanosoma cruzi infection Aparecida Donizette Malvezi a, Rubens Cecchini a, Fausto de Souza a, Carlos Eduardo Tadokoro b, Luiz Vicente Rizzo b,c,d, Phileno Pinge-Filho a
d
a,*
Department of Pathological Sciences, State University of Londrina, CEP:86051-970. Parana, Brazil b Department of Immunology, University of S~ao Paulo, S~ao Paulo, Brazil c F undacß~ao Zerbini, S~ao Paulo, Brazil University of S~ao Paulo Medical School, Div. of Allergy and Clinical Immunology, S~ao Paulo, Brazil Received 14 July 2003; received in revised form 18 November 2003; accepted 12 January 2004 First published online 5 February 2004
Abstract Trypanosoma cruzi infection in mice is associated with severe hematological changes, including anemia, which may contribute to mortality. TNF-a and nitric oxide (NO) play a critical role in establishing host resistance to this pathogen. We hypothesized that phagocyte-derived NO damages erythrocytes and contributes to the anemia observed during T. cruzi infection. To test this hypothesis, two strains of mice that differed in susceptibility and NO response to T. cruzi infection were used in these studies. We also blocked endogenous NO production by aminoguanidine (AG) treatment or blocked TNF-a with a neutralizing antibody and used mice that cannot produce phagocyte-derived NO (C57BL/6 iNOS= ). Following infection with T. cruzi, resistant (C57BL/6) and susceptible (Swiss) mice displayed a parasitemia that peaked at the same time (i.e., day 9), yet parasitemia was 3-fold higher in Swiss mice (P < 0:05). All Swiss mice were dead by day 23 post-infection, while no C57BL/6 mice died during the study. At 14 days postinfection anemia in C57BL/6 mice was more severe than in Swiss mice. Treatment of both strains with the NO inhibitor, AG (50 mg/ kg), and the use of iNOS= mice, revealed that the anemia in T. cruzi-infected mice is not caused by NO. However, the reticulocytosis that occurs during infection was significantly reduced after treatment with AG in both Swiss and C57BL/6 mice (P < 0:05). In addition, we showed that neutralization of TNF-a in vivo induced a significant increase in circulating reticulocytes in T. cruziinfected C57BL/6 mice (P < 0:05), but did not modify other hematologic parameters in these mice. The evaluation of the oxidative stress after induction by t-butyl hydroperoxide (t-BHT) revealed that the treatment with AG completely protected against NOmediated haemoglobin oxidation. Further, treatment with AG, but not with anti-TNF-a, protected against the infection-induced reduction of antioxidant capacity of erythrocytes as assessed by oxygen uptake and induction time. In summary, this is the first report showing the participation of NO and TNF-a in the oxidative stress to erythrocytes in acute T. cruzi infection. Further, our data suggest that NO does not play a direct role in development of the anemia. However, NO may contribute to other hematological changes noted during T. cruzi infection, such as the elevation of circulating reticulocytes and the reduction in circulating leukocytes and neutrophils. 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Trypanosoma cruzi; ChagasÕ disease; Anemia; Oxidative stress; Nitric oxide; TNF-a
1. Introduction *
Corresponding author. Tel.: +55-43-3371-4267; fax: +55-43-33714207. E-mail address: [email protected] (P. Pinge-Filho).
ChagasÕ disease or American trypanosomiasis is caused by the protozoan Trypanosoma (Schizotrypanum) cruzi [1]. The infection is endemic in Latin
0928-8244/$22.00 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsim.2004.01.005
70
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
America and serological data indicate that over 16 million people are infected [2]. In his pioneering work, Carlos Chagas was first to report anemia in patients suffering from a T. cruzi infection [1]. Since the discovery of ChagasÕ disease over 90 years ago, few studies have explored the causes of anemia associated with T. cruzi infection. Mice inoculated with different strains of the T. cruzi present intense thrombocytopenia [3] and neutropenia following by neutrophilia and eosinophilia [4]. Similar hematological alterations have been also described in experimental African trypanosomiasis [5] and is a common characteristic of human immunodeficiency virus infection [6] and malaria [7,8]. Recently, Marcondes and collaborators [9] showed that experimental acute T. cruzi infection is associated with anemia, thrombocytopenia, leukopenia, and bone marrow hypoplasia and that these alterations can be prevented by nifurtimox (an anti-trypanosomal drug) treatment. However, the mechanisms responsible for these hematological alterations are not well understood. Our hypothesis is that cytokines and nitric oxide secreted during the acute phase of a T. cruzi infection [10] are responsible for the hematological changes observed. IFN-c and TNF-a are among the cytokines that play an important role in host defense against T. cruzi [11]. TNF-a can modulate myelopoiesis, erythropoiesis, lymphopoiesis and thrombocytopoiesis [12]. In malaria, TNF-a has been shown to be an important mediator of anemia [13]. However, no such relationship has been established for TNF-a and the anemia associated with T. cruzi infection. Host resistance to T. cruzi is dependent, in part, upon the expression of inducible NO synthase (iNOS) [14] in macrophages. IFN-c and TNF-a play a central role in stimulating iNOS expression and subsequent nitric oxide ( NO) production by macrophages [14]. However, reactive nitrogen species, such as NO, along with reactive oxygen species (ROS) produced by activated macrophages have been implicated as mediators of tissue damage in numerous pathological conditions [15,16]. Erythrocytes are among the tissues that NO can injury [17,18]. For instance, it has been suggested that NO-mediated deformability of erythrocytes plays a role in sepsis and systemic inflammatory response syndrome [19]. Furthermore, NO can cross the erythrocyte membrane and react with SH groups mainly with haemoglobin-cysteine groups [20]. This reaction yields two products: S-nitroso-haemoglobin (SNOHb) and nitrosyl-haemoglobin (HbNO) [21,22]. The nitrosyl complex slowly oxidizes to methaemoglobin (metHb) in presence of oxygen [22], thus measuring metHb can be a useful index of in vivo NO-mediated oxidative stress [23]. NO produced by macrophages reacts very fast with super oxide anion (O 2 ) yielding peroxynitrite (ONOO ), which is a powerful oxidizing and lipid peroxidizing agent [15].
In murine leishmaniasis, there is evidence of enhanced levels of NO in the blood of infected mice and of peroxynitrite associated with their lesions [24]. The induced damage depend on the balance between NO and O 2 . For example, equal fluxes of O 2 and NO can stimulate lipid peroxidation, a high ratio of NO to O 2 can decrease the peroxidation because of the ability of NO to scavenge peroxyl and alkoxyl radicals [15]. Therefore, we predict that the nature and extent of erythrocyte oxidative injury during infection with T. cruzi will depend upon three factors: (1) the induction of iNOS and thus production of NO in response to the infection; (2) the oxidative stress generated outside of erythrocyte, particularly phagocyte-derived O 2 ; (3) the rate of reaction between NO and either haemoglobin or O 2 . Since in T. cruzi infection the concentration of both NO and O 2 are enhanced [25,26], we hypothesized that endogenous NO is responsible for the hematological changes (i.e., anemia) observed in the acute phase of T. cruzi infection. Therefore, in order to test the possible involvement of NO as a part of injury mechanism in experimental ChagasÕ disease, we carried out studies with two strains of mice that differed in iNOS induction following T. cruzi infection. C57BL/6 mice are resistant to T. cruzi and respond to infection with a strong induction of the iNOS gene in phagocytes, while Swiss mice are susceptible to this parasite and have a more modest induction of iNOS following infection. The importance of endogenous NO production on the hematological changes associated with T. cruzi infection was also tested by using C57BL/6iNOS= mice deficient in iNOS and by treating wild-type C57BL/6 mice with aminoguanidine (AG), an inhibitor of iNOS. We measured haemoglobin oxidation and assessed oxidative stress by measuring tbutyl hydroperoxide (t-BHP) induced oxygen uptake and induction time of isolated erythrocytes. The possible involvement of TNF-a in the anemia associated with T. cruzi infection was investigated by treating wild-type C57BL/6 mice with a neutralizing anti-TNF-a antibody. Our results clearly show that NO produced during T. cruzi infection increases erythrocyte oxidative stress, but is not necessary for the development of anemia in acute T. cruzi infection in mice.
2. Materials and methods This study was reviewed and approved by the Internal Scientific Commission and the Bioethics in Research Committee of the Londrina State University, Londrina, Brazil. 2.1. Mice and Trypanosoma cruzi infection Swiss mice were obtained from the breeding colonies of the animal facility of the Center for Biological Sciences
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
at Londrina State University, Londrina, Brazil. C57BL/6 mice and C57BL/6 iNOS= (mice deficient in iNOS) [27], were obtained from the mouse breeding facilities of the Department of Immunology, Institute of Biomedical Sciences, University of S~ ao Paulo, S~ ao Paulo, Brazil. We used 8 to 12 weeks old male mice in all experiments. Mice were maintained under standard conditions in an animal facility within the Department of Pathological Sciences, Center for Biological Sciences, Londrina State University. Commercial rodent diet (Nuvilab-CR1, Nuvital, Campo Mour~ ao, Brazil) and sterilized water were available ad libitum. All procedures with the animals were in accordance with the guidelines of the Brazilian Code for the Use of Laboratory Animals. The Y strain of T. cruzi was kindly supplied by Dr. Paulo Maria Ferreira de Ara ujo (Institute of Biosciences, Campinas State University, Campinas, S~ao Paulo, Brazil). This strain was maintained by weekly intraperitoneally (i.p.) inoculation of Swiss mice with 2 105 blood trypomastigotes forms. Infective blood trypomastigotes were obtained from T. cruzi-infected mice by drawing blood via cardiac puncture following anaesthetization. Motile blood forms were counted and the desired number of parasites (5 · 103 ) was injected intraperitoneally (i.p.) in Swiss, C57BL/6 and C57BL/6 iNOS= mice. Parasitemia was assessed by counting circulating parasites in 5 ll of blood obtained from a tail vein of infected mice. These data were expressed as the number of parasites per milliliter of blood [28]. Parasitemia and survival rates were determined daily, beginning at the 5th day of infection. 2.2. Treatment of infected mice with aminoguanidine During the first 15 days of infection, Swiss and C57BL/6 mice received daily i.p. injections of saline with or without aminoguanidine (Sigma, St. Louis), an inhibitor of inducible nitric oxide synthase (iNOS), diluted in sterile phosphate-buffered saline (50 mg/kg of body weight). The first dose was given 4 h after T. cruzi infection. The dose of AG chosen for these experiments was based on previously published studies demonstrating its efficacy [29].
71
cardiac puncture with heparinized needles and syringes and counted by standard methods [30]. Haemoglobin concentration was determined by the Drabkin method. Hematocrits were obtained by microcentrifugation of capillary tubes filled with heparinized blood [31]. For enumerating reticulocytes, 40 ll of blood heparinized was incubated with 20 ll brilliant cresyl blue for 20 minutes at 37 C, then thin blood smears were prepared on glass slides. After the blood smears air dried, reticulocytes were counted by light microscopy. All blood analysis and cell counts were performed 0, 7 and 14 days post-infection.
2.5. Bone marrow cell harvest Bone marrow cells were harvested by flushing the femoral shafts with ice-cold PBS, as previously described [32]. The total number of cells collected was determined by manual hemacytometer count. For differential counts, cell suspensions from uninfected and day 14-infected mice were deposited on glass slide, stained with May Gr€ unwald-Giemsa, then counted by light microscopy.
2.6. Determination of oxygen uptake and induction time (T ind ) in erythrocytes Heparinized blood samples from uninfected and infected mice (day 14 post-infection) were used for erythrocyte oxidative stress determinations [15,33,34]. After removal of plasma and white cells from whole blood, the remaining erythrocytes were washed twice with a 10 mM sodium phosphate buffered saline (0.9% NaCl, pH 7.4), then resuspended in the same buffer (1:99, v/v). Both t-butyl hydroperoxide 2 mM (t-BHT) induced oxygen uptake and induction time were measured with a Clark-type oxygen electrode at 37 C [33,34]. The induction time is directly related to the intracellular protective antioxidant capacity, while oxygen uptake is an indirect measure of the susceptibility of erythrocytes membranes to lipid peroxidation elicited by t-BHT [15,34].
2.3. In vivo neutralization of TNF-a C57BL/6 mice were injected i.p with 200 lg of an anti-TNF-a mAb (XT22) [10], kindly provided by Dr. Ises A. Abrahamsohn, Department of Immunology, University of S~ ao Paulo, S~ ao Paulo, Brazil. XT22 was given on days 7, 9, 11 and 13 of T. cruzi infection. 2.4. Hematological methods Erythrocytes and leukocytes from normal and infected mice under ether anesthesia, were collected from
2.6.1. Haemoglobin oxidation measurement Oxidation of haemoglobin was determined using a dual beam spectrophotometer (Varian 634S). Erythrocytes were lysed with distilled-deionized water and the spectrum of a 1% solution of lysates was measured from 500 to 700 nm. The signal was digitized and analyzed with the Lynx program (Lynx Tecnologia Eletr^ onica, Ltda). The concentration of oxyhaemoglobin, metahaemoglobin and haemichrome were calculated using the Winterbourn equations [35].
72
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
2.7. Statistical analysis Data are expressed as mean standard deviation of the means. The impact of infection and other treatments were determined by two-way Analysis of Variance (ANOVA). When significant main effects were noted, differences between individual groups were tested using Bonferonni multiple comparisons method. Survival curves were compared using the Mantel-Haenszel log rank test. All statistical analysis was made using GraphPad Prism version 3.0 (GraphPad Software, San Diego, CA).
3. Results 3.1. Course of infection Five and seven days following inoculation with T. cruzi, parasitemia was similar between susceptible (Swiss) and resistant (C57BL/6) strains of mice (Fig. 1(a)). Parasitic load in the blood peaked 9 days post-infection and was significantly higher in Swiss compared to C57BL/6 mice. C57BL/6 iNOS= mice showed parasitemia levels more similar to the susceptible Swiss mice than their parental C57BL/6 strain. There were large and statistically significant differences (P < 0:001) in survival between all three strains of mice studied (Fig. 1(b)). All of the C57BL/6 mice survived the infection, while all of the Swiss and C57BL/6 iNOS= mice died by 23 and 15 days post-infection, respectively. 3.2. Development of anemia We conducted blood cells counts and measured several haematological values in uninfected and T. cruziinfected mice at 7 and 14 days post-infection. Based on significant decreases in haemoglobin, hematocrit and erythrocyte numbers, all three mouse strains studied suffered from anemia 14 days post-infection (Table 1). The severity of the anemia was greatest for C57BL/6 mice. Inhibition of endogenous NO production via AG treatment and neutralizing TNF-a did not impact the extent of anemia that developed 14 days post-infection (Table 1). Both 7 and 14 days post-infection we noted a significant increase in reticulocytes (P < 0:001) present in both the Swiss and C57BL/6 mice (Fig. 2(a) and (b)). Treatment of Swiss mice with AG completely abrogated this increase in reticulocytes associated with T. cruziinfection. Similarly, AG treatment of C57BL/6 mice significantly reduced the reticulocytosis associated with T. cruzi-infection in this resistant strain of mice. In the C57BL/6 iNOS= mice we noted a transient reticulocytosis, with a significant elevation in reticulocytes at 7, but not at 14 days post-infection (Fig. 2(b)). Surprisingly, we observed a very large increase in reticulocytes
Fig. 1. Parasitemia (a) and survival (b) of mice infected with T. cruzi. Swiss, C57BL/6 and C57BL/6 iNOS= mice were infected with 5 103 blood trypomastigotes Y strain T. cruzi. Parasitemia was quantified as trypomastigotes per milliliter of blood. Results are expressed as the mean standard deviation from of 10 mice per group, in an experiment representative of three similar experiments. Results were analysed by analysis of variance (ANOVA) followed by DunnÕs nonparametric test. * P < 0:05, significant difference in parasitemia, C57BL/6 and C57BL/6 KO iNOS= vs Swiss. ** Survival curve significantly different from the of C57BL/6 mice, P < 0:01.
in the T. cruzi-infected mice treated with the neutralizing antibody against TNF-a (Fig. 2(b); P < 0:001). 3.3. Changes in blood leukocytes and neutrophils Fourteen days following T. cruzi-infection, Swiss mice presented with neutropenia (50% decline; P < 0:05) without changes in mononuclear leukocyte numbers, while C57BL/6 mice showed a 50% decline in leukocytes (P < 0:05) and no change in neutrophils (P > 0:05) (Table 1). Treatment of C57BL/6 mice with AG tended to diminish the decline in leukocyte numbers associated with infection. In contrast, AG treatment in Swiss mice significantly reduced leukocyte numbers (25–30%; P < 0:05) in uninfected and infected mice alike. Treatment of Swiss mice also modestly reduced blood neutrophils in uninfected mice (P < 0:05), but it was able to partially restore neutrophil counts following T. cruzi-infection (P < 0:05). Neutralization of endoge-
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
73
Table 1 Hematological values on day 14 in T. cruzi-infected micea Group
Haemoglobin (g/dl)
Hematocrit (%)
Erythrocytes (109 )
Leukocytes (106 )
Neutrophils (106 )
Swiss C C + AG I I + AG
13.8 0.5 13.7 1.3 10.8 0.9 10.3 0.6
44.9 1.3 42.6 0.5 35.5 3.8 32.4 2.0
7.0 0.5 7.9 0.2 6.4 0.5 6.1 0.5
5.4 0.7 4.0 0.6 5.4 1.5 3.7 0.5
1.2 0.3 0.9 0.2 0.5 0.3 0.9 0.2
C57BL/6 C C + AG I I + AG I + Anti-TNF
13.5 0.7 12.6 0.9 8.2 0.7 8.6 0.9 8.7 0.7
44.0 0.8 40.3 3.5 28.1 1.4 28.3 2.6 28.3 2.3
7.6 0.8 7.0 0.3 5.1 0.5 5.3 0.3 5.0 0.3
5.0 0.7 5.4 1.2 3.0 0.8 4.0 0.7 5.0 0.8
0.4 0.02 0.5 0.1 0.4 0.1 1.1 0.1 1.3 02
INOS-/C I
12.7 1.1 9.3 1.2
40.5 3.4 29.8 2.4
7.4 0.5 5.2 0.6
5.4 1.2 2.6 0.7
1.6 0.2 0.4 0.1
a Groups of mice were infected (I) with 5 103 T. cruzi or infected and treated with aminoguanidine (AG) or anti-TNF-a mAb. (c) control; (C + AG) Control and AG-treated; (I) infected; (I + AG) infected and treated with AG. Values represent the mean SD and are representative of two independent experiments, using 5–12 mice per group. * Significantly different (P < 0:05) when compared with non-infected group. ** Significantly different (P < 0:001) when compared with non-infected group.
marked leukopenia and neutropenia (i.e., a 50% and 75% decline, respectively; P < 0:05). 3.4. Changes in cellularity and erythroblasts in the bone marrow
Fig. 2. Percentage of reticulocytes in the blood. Swiss mice (a), C57BL/ 6 and iNOS= mice (b) were infected with 5 103 T. cruzi. Four hours following infection and then daily for 15 days some of the mice in each strain received an i.p. injection of aminoguanidine (AG, 50 mg/kg BW) or 200 lg of anti-TNFa mAB on days 7, 9, 11 and 13 post-challenge. Control mice received saline injections only. Values represent the mean standard deviation and are representative of two independent experiments, with 5–12 mice per group. Differences in percentage of reticulocytes were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, P < 0:05.
nous TNF-a was able to completely inhibit the decline in blood leukocytes associated with T. cruzi-infection in C57BL/6 mice. The C57BL/6 iNOS= mice showed a
T. cruzi infection in Swiss and C57BL/6 mice was associated with a significant decrease in total number of bone marrow cells, in particular erythroblasts (Fig. 3(a)– (d); P < 0:05). AG treatment had no significant effect on the hypoplasia in the bone marrow of Swiss mice (Fig. 3(a) and (c)). In contrast, AG treatment of C57BL/ 6 mice prevented some of the decline in bone marrow cellularity associated with T. cruzi infection (Fig. 3(b) and (d); P < 0:001). However, the T. cruzi infectioninduced decline in total cells and erythroblasts in the bone marrow of C57BL/6 iNOS= was nearly identical to that observed in C57BL/6 mice (Fig. 3(b) and (d)). Neutralization of endogenous TNF-a in C57BL/6 mice prevented the decline in erythoblasts without significantly changing the decline in total cellularity in the bone marrow of infected mice (Fig. 3(d)). 3.5. Erythrocyte oxidative stress The ratio of oxyhaemoglobin to methaemoglobin (i.e., % oxidation) was increased following T. cruzi infection in both Swiss and C57BL/6 mice (Table 2). The extent of oxidation tended to be higher in Swiss compared to C57BL/6 mice. As expected, treatment with AG completely protected against this NO-mediated oxidation of haemoglobin. We noted an increase in the % oxidation of haemoglobin in C57BL/6 iNOS= mice, this appeared to be primarily a consequence of the
74
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
Fig. 3. Bone marrow cells and erythroblasts in T. cruzi-infected mice. Swiss mice (a), C57BL/6 and iNOS= mice (b) were infected with 5 103 T. cruzi. Four hours following infection and then daily for 15 days some of the mice in each strain received an i.p. injection of aminoguanidine (AG, 50 mg/kg BW) or 200 lg of anti-TNFa mAB on days 7, 9, 11 and 13 post-challenge. Control mice received saline injections only. Values represent the mean standard deviation and are representative of two independent experiments, with 5–12 mice per group. Differences in percentage of reticulocytes were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, P < 0:05. Table 2 Oxyhaemoglobin (OxyHb) oxidation ratea in T. cruzi-infected miceb Group
OxyHb (lM)
MetaHb + Hemic. (lM)
Rate oxidation (%)
Swiss C C + AG I I + AG
19.5 22.0 17.7 19.4
5.6 6.8 8.7 4.5
28.7 30.1 49.1 23.1
C57BL/6 C C + AG I I + AG I + Anti-TNF
24.0 20.8 16.3 18.8 14.8
5.2 3.0 5.2 4.6 4.8
21.6 14.4 31.9 24.4 32.4
INOS-/C I
29.2 17.9
8.2 6.7
28.0 37.4
a Oxyhaemoglobin (OxyHb) oxidation rate on day 14 of T. cruzi infection. b Groups of mice were infected (I) with 5 103 T. cruzi or infected and treated with aminoguanidine (AG) or anti-TNF-a mAb. (c) control; (C + AG) Control and AG treated; (I) infected; (I + AG) infected and treated with AG. Values are representative of two independent experiments, using 5–12 mice per group.
infection-induced anemia with the associated decline in haemoglobin levels. Anti-TNF-a treatment did not affect haemoglobin oxidation in C57BL/6 mice infected with T. cruzi. Oxygen uptake by erythrocytes is directly associated with the susceptibility of the erythrocyte membrane to undergo lipid peroxidation elicited by t-BHP and is
proportional to previous oxidative stress experienced by the erythrocyte in vivo. The oxygen uptake by erythrocytes from Swiss mice was significantly increased 14 days following infection (Fig. 4(a); P < 0:05). AG treatment significantly diminished erythrocyte oxygen uptake in both uninfected and infected Swiss mice (Fig. 4(a); P < 0:001). In C57BL/6 mice T. cruzi infection was associated with a significant decrease in oxygen uptake at 14 days post-infection (Fig. 4(b); P < 0:001). AG treatment enhanced oxygen uptake in uninfected mice, but diminished it following infection. Prior to infection, iNOS= mice showed very low levels of oxygen uptake, which increased dramatically following infection. Treatment with anti-TNF-a did not alter the oxygen uptake profile of erythrocytes from C57BL/6 mice prior to or after T. cruzi infection. Induction time (Tind ) is directly related to the intracellular protective antioxidant capacity of the erythrocyte. Neither infection nor AG treatment altered Tind in Swiss mice (P > 0:05, data not shown). In contrast, T. cruzi infection in the C57BL/6 mice resulted in a significant reduction in Tind (Fig. 5; P < 0:05). Treatment with AG completely abrogated the infection associated decline in Tind (P < 0:01, Fig. 5). Prior to infection, Tind was significantly longer in erythrocytes isolated from C57BL/6 iNOS= mice compared to C57BL/6 mice. However, like wild-type C57BL/6 mice, a significant decrease in Tind was seen for the C57BL/6 iNOS= 14 days post-infection (P < 0:001). Treatment with antiTNF-a also protected against the reduction of antioxidant capacity of C57BL/6 erythrocytes on day 14 post-infection (Fig. 5; P < 0:001).
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
Fig. 4. Erythrocyte oxidative stress during T. cruzi infection. Here we measured oxygen uptake induced by t-BHT in erythrocyte from Swiss mice (a) and from C57BL/6 and iNOS= mice (b). Four hours following infection and then daily for 15 days some of the mice from strain received an i.p. injection of aminoguanidine (AG, 50 mg/kg BW) or 200 lg of anti-TNFa mAb on day 7, 9, 11 and 13 post-challenge. Controls received saline injections only. Values represent the mean standard deviation and are representative of two independent experiments, using 5 mice per group. Differences in oxygen uptake were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, *P < 0:05.
Fig. 5. Erythrocyte antioxidant capacity following T. cruzi infection. We measured induction time (Tind ) following t-BHT treatment of erythrocytes from C57BL/6 and C57BL/6 iNOS= mice. Mice were treated as described in the previous figure. Values represent the mean standard deviation and are representative of two independent experiments, using 5 mice per group. Differences in induction time levels were analyzed by two-way analysis of variance (ANOVA) followed by BonferroniÕs test. Means not sharing a letter differ, *P < 0:05.
4. Discussion Infection of mice with T. cruzi is characterized by exponential growth of parasites and anemia in both resistant and susceptible strains of mice. In this study we demonstrated that T. cruzi infection is lethal in susceptible mouse strains (i.e., Swiss and C57BL/6 iNOS=
75
mice). Similar results have been reported previously [3,36]. Resistance to T. cruzi infection in mice depends on the production of TNF-a and NO [37]. However, excessive NO production may result in severe tissue damage, including erythrocyte destruction. Acute infection with T. cruzi has been associated with significant alterations in hematological values in the blood and bone marrow [9]. We showed that compared to susceptible mice, T. cruzi infection in resistant mice resulted in a more severe anemia (i.e., reduced Hb and Hct). Infection was also associated with a time-dependent increase in blood reticulocytes and a decrease in bone marrow erythroblasts (two markers of anemia). Consistent with our Hb and Hct data, the changes in reticulocytes and erythroblasts tended to be greater in resistant (C57BL/6) versus susceptible mice (Swiss). The primary objective of this study was to determine if endogenous NO production was responsible for the hematological changes (i.e., anemia) observed in a T. cruzi infection. Blocking endogenous NO production via AG treatment had no discernable impact on the infection-induced changes in hematological parameters measured in this study for either susceptible or resistant mouse strains. We did not directly measure in vivo NO production, however, the ability of AG treatment to abrogate NO-mediated oxidation of haemoglobin during the infection supports the conclusion that AG was effectively inhibiting NO production in vivo. The occurrence of anemia in C57BL/6 iNOS= mice following infection with T. cruzi provides further evidence that NO is not required for the development of infectioninduced anemia. However, our data clearly demonstrate that endogenous NO production is essential for the control of parasitemia and survival following a T. cruzi infection. Similar results have been reported previously [36]. AG treatment was able to modify two indirect indicators of anemia. First, AG treatment prevented the T. cruzi infection-induced increase in blood reticulocytes in both susceptible and resistant strains of mice. Second, the decrease in bone marrow erythroblasts that occurred 14 days post-infection was not as large if mice were treated with AG. Also, AG treatment was able to prevent neutropenia in Swiss mice and diminish the leukopenia in C57BL/6 mice at day 14 post-infection, probably due a reduction in the cellular toxicity provoked by NO. In addition, we showed that in vivo neutralization of TNF-a with monoclonal antibody (XT22) treatment, provoked a dramatic increase in the percentage of reticulocytes in C57BL/6 mice 14 days after T. cruzi infection, but didnÕt modify the anemia of those animals. The leukopenia observed in infected C57BL/6 mice was abolished for the treatment with anti-TNF-a. These data suggest that the TNF-a produced by activated macrophages during the acute infection with the T. cruzi
76
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77
diminishes erythropoiesis. In fact, previous studies have demonstrated inhibitory effects of TNF-a on erythropoiesis [12,38]. In this study we assessed oxidative stress in erythrocytes three ways: by measuring oxyhaemoglobin and haemoglobin derivatives (i.e., methaemoglobin and hemichrome), oxygen uptake and induction time (Tind ) following treatment with t-butyl hydroperoxide, t-BHT. We found that an acute infection of mice with T. cruzi was generally associated with increased erythrocyte oxidative stress. For example, we observed a modest, but significant increase the ratio of metHb and hemichrome over oxyHb following T. cruzi infection in Swiss and C57BL/6 mice. Further, we found evidence of lipid peroxidation in erythrocytes (i.e., increased oxygen uptake) from T. cruzi-infected Swiss mice. That AG treatment reduced erythrocyte oxygen uptake both prior to and following infection suggested that NO played a role in lipid peroxidation of erythrocyte membranes. Surprisingly, and in contrast to our observations in Swiss mice, T. cruzi infection in C57BL/6 mice significantly reduced in vitro oxygen uptake by erythrocytes. However, blocking in endogenous NO production in C57BL/6 mice by AG treatment significantly reduced oxygen uptake by erythrocytes just as it had done in Swiss mice. We do not know for certain how infection causes oxidative stress in erythrocytes, but we believe that phagocyte-derived NO and O are central to this 2 process. O can react with NO yielding ONOO or by 2 the Fenton reaction produce OH, both of which could lead to oxidative stress in erythrocytes. Since Swiss mice are only modestly iNOS responsive, in vivo NO formation following T. cruzi infection is expected to be lower is Swiss than C57BL/6 mice. With only modest NO production, we believe that a larger fraction of NO in Swiss mice will be present as HbFe(II)(NOþ ) and consequently lead to higher relative yields of nitrosylhaemoglobin (SNOHb) [22]. The nitrosyl-haemoglobin complex slowly oxidizes to metHb, which is known to reduce the flux of O 2 [22]. The net result should be lower internal oxidative stress in erythrocytes from Swiss than C57BL/6 mice. Our observation that antioxidant capacity as evaluated by Tind was not significantly affected by T. cruzi infection in Swiss mice is consistent with this theoretical mechanism. Our results obtained with C57BL/6 mice suggested that the nature of the oxidative stress that occurred during an acute T. cruzi infection was somewhat different than that found in erythrocytes from Swiss mice. In particular, the greater iNOS induction seemed to contribute to a higher oxidation rate of Hb than we observed in Swiss mice. However, similar to Swiss mice, prior in vivo AG treatment of C57BL/6 mice abrogated most of this infection-induced Hb oxidation and diminished erythrocyte oxygen uptake as well. Unlike in
Swiss mice, our data in C57BL/6 mice suggest that oxidation of Hb to metHb does diminish the antioxidant defense capacity of erythrocytes. It has been shown that besides intracellular GSH, HbSH groups contribute to the antioxidant capacity of erythrocytes [23]. These data suggests that when NO is abundant intracellular oxidative stress is favored over oxidative attack of the erythrocyte membrane. In fact in the complete absence of NO, membrane injury as assessed by oxygen uptake was greatly increased in erythrocytes isolated from C57BL/6 iNOS= 14 days after infection. These results suggest that participation of NO in membrane oxidative injury may be indirect. Altogether our results indicate that in experimental ChagasÕ disease oxidative stress inside and outside of the erythrocyte occurs, but that the precise nature of this oxidative stress may depend on the balance of NO/O 2 produced by the host. We conclude that death associated with an acute T. cruzi infection is not caused by NO-induced anemia or alterations in bone marrow cellularity. Our data confirm the importance of NO as part of the hostÕs defense against T. cruzi infection. However, the impact of NO on cells of the blood and bone marrow appears to be complex, and NO-based therapies against infections are not easy to design and demand caution.
Acknowledgements We thank Vargas, J.A. for excellent technical assistance, Araujo, P.M.F (State University of Campinas) for providing T. cruzi (Y strain). We are also grateful to Abrahamsohn, I.A for anti-TNF-a mAb (University of S~ao Paulo) and Dr. Kevin L. Fritsche (University of Missouri) for his assistance with editing the manuscript. Fausto de Souza was the recipient of a CNPq/PIBIC/ UEL undergraduate fellowship.
References [1] WHO – World Health Organization (1997). Tropical disease research: progress report 95–96. p. 134. [2] Chagas, C. (1909) Nova tripanozomiase humana. Estudos sobre a morfologia e o ciclo evolutivo do Schizotrypanum cruzi n.gen., s.sp. agente etiol ogico de uma nova entidade m orbida do homem. Mem. Inst. Oswaldo Cruz 1, 159–218. [3] Cardoso, J.E. and Brener, Z. (1980) Hematological changes in mice experimentally infected with Trypanosoma cruzi Mem. Inst. Oswaldo Cruz 75, 97–104. [4] Repka, D., Rangel, H.A., Atta, A.M., Gavino, V.A. and Piedrabuena, A.E. (1985) Experimental Chagasas diseases in mice infected with one LD50 of parasite. Rev. Bras. Biol. 45, 309–316. [5] Ikede, B.O., Lule, M. and Terry, R.J. (1977) Anemia in trypanosomiasis: mechanisms of erythrocyte destruction in mice infected with Trypanosoma congolense or T. bruce i. Acta Trop. 34, 53–60.
A. Donizette Malvezi et al. / FEMS Immunology and Medical Microbiology 41 (2004) 69–77 [6] Claster, S. (2002) Biology of anemia, differential diagnosis, and treatment options in human immunodeficiency virus infection. J. Infect. Dis. 85, 105–109. [7] Weatherall, D.J., Miller, L.H., Baruch, D.I., Marsh, K., Doumbo, O.K., Casals-Pascual, C. and Roberts, D.J. (2002) Malaria and red cell. Hematology (Am. Soc. Hematol. Educ. Program), 35–57. [8] Paul, R.E. and Brey, P.T. (2003) Malaria Parasites and red blood cells: from anaemia to transmission. Mol. Cells 15, 139–149. [9] Marcondes, M.C., Borelli, P., Yoshida, N. and Russo, M. (2000) Acute Trypanosoma cruzi infection is associated with anemia, thrombocytopenia, leukopenia, and bone marrow hypoplasia: reversal by nifurtimox treatment. Microbes Infect. 4, 347–352. [10] Abrahamsohn, I.A. and Coffman, R.L. (1995) Cytokine and nitric oxide regulation of the immunosuppression in Trypanosoma cruzi infection. J. Immunol. 155, 3955–3963. [11] Munoz-Fernandez, M.A., Fernandez, M.A. and Fresno, M. (1992) Synergism between tumor necrosis factor-alpha and interferon-gamma on macrophage activation for the killing of intracellular Trypanosoma cruzi through a nitric oxide-dependent mechanism. Eur. J. Immunol. 22, 301–307. [12] Ulich, T.R., Shin, S.S. and del Castillo, J. (1993) Haematologic effects of TNF. J. Res. Immunol. 144, 347–354. [13] Miller, K.L., Silverman, P.H., Kullgren, B. and Mahlmann, L.J. (1989) Tumor necrosis factor alpha and the anemia associated with murine malaria. Inf. Immun. 5, 1542–1546. [14] MacMicking, J., Xie, Q.W. and Nathan, C. (1997) Nitric oxide and macrophage function. Annu. Rev. Immunol. 5, 323–350. [15] Halliwell, B. and Gutteridge, J.M. (1999) Free Radicals, Other Reactive Species and Disease. In: Free radicals in biology and medicine, pp. 660–699. Oxford University Press. [16] Oliveira, F.J.D.A. and Cecchini, R. (2000) Oxidative stress of liver in hamsters infected with Leishmania (L) chagasi. J. Parasitol. 86, 1067–1072. [17] Biswas, T., Pal, J.K., Naskar, K., Ghosh, D.K. and Ghosal, J. (1995) Lipid peroxidation of erythrocytes during anemia of the hamsters infected with Leishmania donovani. Mol. Cel. Bioch. 146, 99–105. [18] Biswas, T., Ghosh, D.K., Mukherjee, N. and Ghosal, J. (1997) Lipid peroxidation of erythrocytes in visceral leishmaniasis. J. Parasitol. 83, 151–152. [19] Baterman, R.M., Jagger, J.E., Sharpe, M.D., Ellsworth, M.L., Mehta, S. and Ellis, C.G. (2001) Erythrocyte deformability is a nitric oxide-mediated factor in decreased capillary density during sepsis. Am. J. Physiol. Heart Circ. Physiol. 280, 2848–2856. [20] Baskurt, O.K., Temiz, A. and Meiselman, H.J. (1998) Effect of superoxide anions on red blood cell rheologic properties. Free Radic. Biol. Med. 24, 102–110. [21] Rassaf, T., Bryan, N.S., Maloney, R.E., Speciean, V., Kelm, M., Kalyanaraman, B., Rodrigues, J. and Feelisch, M. (2003) NO adducts in mammalian red blood cells: too much or too little. Nat. Med. 9, 481–483. [22] Herold, S. and Rock, G. (2003) Reactions of Deoxy-Oxy-, and Methaemoglobin with Nitrogen Monoxide. J. Biol. Chem. 278, 6623–6634. [23] Rossi, R., Barra, D.B.A., Boumis, G., Canofeni, S., Simplicio, P.D., Lusini, L., Pascarella, S. and Amiconi, G. (1998) Fastreacting thiols in rat haemoglobins can intercept damaging species in erythrocytes more efficiently than glutathione. J. Biological Chemistry 273, 19198–19206.
77
[24] Augusto, O., Linares, E. and Giorgio, S. (1996) Possible roles nitric oxide and peroxynitrite in murine leishmaniasis. Bras. J. Med. Biol. Res. 29, 853–862. [25] Russo, M., Starombinas, N., Ribeiro-Dos-Santos, R., Minoprio, P. and Hontebeyrie-Joskowicz, M. (1989) Suscetible mice present higher macrophages activation than resistant mice during infections with myotropc strains of Trypanosoma cruzi. Parasite Immunol. 11, 385–395. [26] Cardoni, R.L., Rottemberg, M.E. and Segura, E.L. (1990) Increased production of reactive oxygen species by cells from mice acutely infected with Trypanosoma cruzi. Cell. Immunol. 1, 11–21. [27] Macmicking, J.D., Nathan, C., Hom, G., Chartrain, N., Fletcher, D.S., Trumbauer, M., Stevens, K., Xie, Q., Sokol, K., Hutchinson, N., Chen, H. and Mudgett, J.S. (1995) Altered responses to bacterial infection and endotoxic shock in mice lacking inducible Nitric oxide synthase. Cell 81, 641– 650. [28] Brener, Z. (1962) Therapeutic activity and criterion from cure on mice experimentally infected with Trypanosoma cruzi. Rev. Inst. Med. Trop. S~ao Paulo 4, 389–396. [29] Martins, G.A., Vieira, L.Q., Cunha, F.Q. and Silva, J.S. (1999) Gamma interferon modulates CD95 Fas and CD95 ligand (Fas-L) expression and nitric oxide-induced apoptosis during the acute phase of Trypanosoma cruzi infection: a possible role in immune response control. Infect. Immun. 67, 3864–3871. [30] Dacie, J.V. and Lewis, S.M. (1994) Basic haematological techniques. In: Practical Haematology (Dacie, J.V. and Lewis, S.M., Eds.), eighth ed. Churchill Livingstone, Edinburg. [31] Bain, B.J. (1997). Celulas Sanguıneas: Um guia pratico, 2nd ed., pp 25-51. Porto Alegre, Artes Medicas Sul LTDA. [32] Tadokoro, C.E. and Abrahamsohn, I.A. (2001) Bone marrowderived macrophages grown in GM-CSF or M-CSF differ in their ability to produce IL-12 and to induce IFN-gamma production after stimulation with Trypanosoma cruzi antigens. Immunol. Lett. 77, 31–38. [33] Lissi, E.R., Fransj, C., Fernandes, V. and Videla, L.A. (1986) Effects of antioxidants and haemoglobin status on the t-butyl hydeoperoxide-induced oxygen uptake by red blood cells. Cell. Biochem. Funct. 4, 447–454. [34] Bainy, A.C.D., Saito, E., Carvalho, P.S.M. and Junqueira, V.B.C. (1996) Oxidative stress in gill, erythrocytes, liver and kidney of Nile tilapia (Oreochromis niloticus) from a polluted site. Aquatic Toxicol. 34, 151–162. [35] Winterbourn, C.C. (1985). Reaction of superoxide with haemoglobin. CRC handbook of methods for oxigen radical research, pp 137–141. [36] Saeftel, M., Fleischer, B. and Hoerauf, A. (2001) Stage-dependent role of nitric oxide in control of Trypanosoma cruzi infection. Infect. Immun. 69, 2252–2259. [37] Silva, S.S., Vespa, G.N.R., Cardoso, M.A.G., Aliberti, J.C.S. and Cunha, F.Q. (1995) Tumor necrosis factor alfa mediates resistance to Trypanosom cruzi infection in mice by inducing nitric oxide production in infected gamma interferon-activated macrophages. Inf. Immun. 12, 4862–4867. [38] Ulich, T.R., Castillo, J. and Yin, S. (1990) Tumor necrosis factor exerts dose-dependent effects on erythropoiesis and myelopoiesis in vivo. Exp. Hematol. 18, 311–315.