in-vivo treatment with benznidazole enhances phagocytosis, parasite destruction and cytokine release...

In-vivo treatment with benznidazole enhances phagocytosis, parasite

destruction and cytokine release by macrophages during infection

with a drug-susceptible but not with a derived drug-resistant

Trypanosoma cruzi population

S.M.F.MURTA1,3, C.ROPERT2, R.O.ALVES3, R.T.GAZZINELLI1,2 & A.J.ROMANHA3

1Departamento de Bioquı´mica e Imunologia, ICB-UFMG, 30270–010 BH, MG, Brazil,2Laboratorio de Doenc¸a de Chagas and3Laboratorio de Parasitologia Celular e Molecular, Centro de Pesquisas Rene´ Rachou, FIOCRUZ, 30190–002 BH, MG, Brazil

SUMMARY

To study the effect of chemotherapy on parasite–macrophageinteraction we used the wild-type Y strain (drug-susceptible)of Trypanosoma cruziand a drug-resistant parasite popula-tion derived from the same strain. Trypomastigotes isolatedfrom untreated infected mice, as well as, 3 h after treatmentwith BZ were incubated with inflammatory macrophagesand used to study phagocytosis, parasite destruction, cytokinerelease and reactive nitrogen intermediates (RNI) synthesis.Phagocytosis and destruction of the drug-susceptible para-sites were significantly enhanced by drug treatment. Theseenhancements were accompanied by an increase in cytokines[interleukin (IL)-12 and tumour necrosis factor (TNF)a] andRNI release by murine inflammatory macrophages primedwith IFN-g. In contrast, BZ treatment of mice infected withdrug-resistantT. cruzipopulation showed no effect whatso-ever. The synthesis of IFN-g and RNI by splenocytes of miceinfected with either susceptible and drug-resistant parasitepopulations, before and after treatment with BZ were alsostudied. Only the splenocytes from mice infected with thedrug-susceptible parasites treated with BZ produced highlevels of IFN-g and RNI. Our findings indicate that BZ acts onthe drug-susceptibleT. cruzi parasites by enhancing thephagocytosis and the production of cytokines and RNI,thus, favouring the destruction of the intracellular parasitesby the cellular compartment of the immune system.

Keywords Trypanosoma cruzi,Chagas’ disease, drugresistance, nitric oxide, interferon gamma, interleukin-12,tumour necrosis factor alpha

INTRODUCTION

Chagas’ disease, caused byTrypanosoma cruziaffectsapproximately 18 million inhabitants on the Americancontinent (World Health Organization 1994). Only twonitroheterocyclic drugs, the nitrofuran nifurtimox and the2-nitroimidazole benznidazole (BZ), are in clinical use atpresent. Both drugs reduce the duration and clinical severityof acute and congenitalT. cruzi infection, but they areeffective in only 50% of treated patients (Khaw & Panosian1995). In addition, these drugs exhibit poor efficacy in thetreatment of the chronic phase of the disease. The existenceof T. cruzistrains naturally resistant to both drugs may be animportant factor in explaining the low rates of cure detectedin treated chagasic patients (Filardi & Brener 1987, Murtaet al. 1998).

The mechanism of action of nitroheterocyclic derivativesagainstT. cruzi is poorly understood. Interestingly, a recentstudy performed in our laboratory (Michailowskyet al. 1998)suggest that activation of the immune system by interleukin(IL)-12 may enhance the efficacy of BZ treatment duringexperimental Chagas’ disease. In addition, phagocytosis andintracellular destruction, by mouse peritoneal macrophage,of blood forms of wild-type Y strain collected from miceafter BZ treatment is significantly enhanced as comparedwith parasites obtained from untreated animals (Lages-Silvaet al. 1990). Thus, the higher efficacy of nitroheterocyclicderivatives during acute infection withT. cruzi, may be inpart explained by the early activation of the immune system.

In fact, the resistance toT. cruziduring acute phase infec-tion is critically dependent on cytokine-mediated macrophageactivation. Interferon (IFN)-g is produced after infectionand has been shown to protect susceptible mice fromT. cruziinfection (Reed 1988, Torricoet al. 1991, Silvaet al. 1992).T. cruzi trypomastigotes induce macrophages to produce

Parasite Immunology, 1999:21: 535–544

q 1999 Blackwell Science Ltd 535

Correspondence: A.J.RomanhaReceived: 5 October 1998Accepted for publication: 29 April 1999

IL-12 and tumour necrosis factor (TNF)a (Silva et al. 1995,Alibert et al. 1996), that triggers the synthesis of IFN-g bynatural killer cells and T lymphocytes (Gazzinelliet al.1993, Cardilloet al. 1996). Both in-vivo and in-vitro experi-ments suggest that activation ofT. cruzi-infected mousemacrophages with IFN-g and TNFa leads to induction ofnitric oxide (NO) synthesis, which mediates trypanostaticand/or trypanocidal functions (Gazzinelliet al. 1992, Vespaet al. 1994, Hunteret al. 1996).

To further investigate the cooperative effect of BZ treat-ment and macrophage activation during therapy of acutephase of Chagas’ disease, we used a previously selected BZresistant population ofT. cruzi (Murta & Romanha 1998)and its parental Y strain (susceptible). Mice infected witheither parasites were treated at the peak of parasitemia witha high dose of BZ. Trypomastigotes isolated from untreatedinfected mice, as well as, 3 h after treatment with BZ wereused to study the parasite–macrophage interactionin vitro.Our results demonstrate that the effector functions were allaugmented when susceptible parasites recovered from treatedanimals were used. Consistent with these observations, ourresults demonstrate that in-vivo synthesis of IFN-g and NOby splenocytes of mice infected with susceptible Y strainwas dramatically enhanced after treatment with BZ. Incontrast, treatment of animals infected with the drug-resistantpopulation had no effect on in-vitro or in-vivo parasite-induced macrophage activation byT. cruzi. Thus, this studyusing susceptible and drug-resistantT. cruzi populationsfurther corroborates the importance of the parasite–immunesystem interaction on the efficacy of BZ therapy duringexperimental Chagas’ disease. It also provides an importantmodel to elucidate specific parasite genotype/phenotypewhich is responsible for resistance to drug treatmentin vivo.

MATERIALS AND METHODS

T. cruzi populations and clones

Populations and clones ofT. cruzi, Y strain (Pereira da Silva& Nussenzweig 1953), WT and BZR, previously selectedinvivo (Murta & Romanha 1998) were used in this study. Theclones used here were: 1, 3, 4, 6, 9, 10, 11, 13, 21 and 25susceptible and 8, 12, 14, 16, 20, 22, 26, 27, 28 and 30resistant to BZ.

Molecular characterization of WT and BZR T. cruzipopulations and clones

Parental and 10 clones of either populations ofT. cruziwere characterized using the following molecular markers:(i) isoenzyme patterns of six enzymes; (ii) genetic variabilityassayed by Randomly Amplified Polymorphic DNA

(RAPD) with three different primers; and (iii) gene probesfor P-glycoprotein (TcPGP) and hypoxanthine-guaninephosphoribosyltransferase (HGPRT). All markers were pre-viously used for molecular characterization of susceptibleand naturally resistant strains ofT. cruzito BZ and nifurtimox(Murta et al. 1998). The parasites were grown in LITmedium (Camargo 1964) at 288C, washed three times inPBS (132 mM NaCl, 3 mM KCl, 8 mM Na2HPO4, 1·5 mMKH2PO4, pH 7·3) by centrifugation for 10 min at 1500g,48C and the parasite pellets were stored at¹ 708C until usedfor the preparation of enzymatic extract and DNA. Iso-enzyme patterns were obtained as described (Carneiroet al.1990). The following six enzymes were analysed: alanineaminotransferase (ALAT) [E.C.2·6.1·2]; aspartate amino-transferase (ASAT) [E.C.2·6.1·1]; glucose phosphate iso-merase (GPI) [E.C.5·3.1·9]; phosphoglucomutase (PGM)[E.C.2·7.5·1]; glucose-6-phosphate dehydrogenase (G6 PD)[E.C.1·1.1·49] and malic enzyme (ME) [E.C.1·1.1·40]. Para-site DNA isolation, amplification, electrophoresis and stain-ing was essentially the same as described elsewhere (Steindelet al. 1993). For RAPD, the three random primers 3302 (50-CTGATGCTAC-30), 3303 (50-TCACGATGCA-30) or 3307(50-AGTGCTACGC-30) were used. To study polymorph-isms, DNA from all clones were subjected to polymerasechain reaction (PCR) with primers specific toT. cruziTcPGPand HGPRT as previously described (Murtaet al. 1998).

Mouse infection

Blood of mice infected with the WT or BZRT. cruzipopulations and clones, was collected from orbital venoussinus (0·3–0·5 ml), diluted in 3·8% sodium citrate (1 : 3 v/v)and inoculated intraperitoneally (i.p.) in normal mice. Groupsof 10 male Swiss albino mice of 18–20 g, were inoculatedi.p. with 104 bloodstream forms.

Treatment of infected mice

Mice at the peak of parasitemia, the 7th day of infection,were treated with a single high dose of 500 mg of BZ/kgof body weight (Filardi & Brener 1984). The BZ (N-benzyl-2-nitro-1-imidazolacetamide), commercially available asRochagan (Roche, SP, Brazil), was dissolved in distilledwater and given to mice by oral route through gavage. Theparasitemia of animals treated and untreated was evaluatedat different times after treatment, as previously described(Brener 1962).

Isolation of T. cruzi bloodstream trypomastigotes

Three hours after drug administration, blood of infectedmice treated and untreated were collected from orbital

S.M.F.Murtaet al. Parasite Immunology

q 1999 Blackwell Science Ltd,Parasite Immunology, 21, 535–544536

venous sinus and defibrinated with glass beads. The bloodwas diluted (v/v) with Dulbecco’s Modified EssentialMedium (DMEM) supplemented with 2 mM L-glutamine,80mg/ml gentamicin, 10% heat-inactivated foetal calf serum(FCS), and centrifuged at 100g for 10 min at room tem-perature. After incubation at 288C for 30 min, the parasitesfrom the supernatant were collected and washed twice withDMEM 10% FCS at 48C, 1000g for 10 min.

Phagocytosis assay

Inflammatory macrophages were obtained from male Swissalbino mice inoculated i.p. with 1·5 ml of 3% sodiumthioglycolate (Difco Laboratories, Detroit, MI, USA). Onthe fourth day of inoculation mouse macrophages wereharvested from peritoneal cavity with DMEM withoutFCS. For cell adhesion 2·2×105 macrophages were usedto seed 5×15 mm glass cover slips in DMEM without FCSand incubated at 378C for 40 min. Non-adherent cells wereremoved by washing and macrophages were cultivated at378C, 5% CO2 in DMEM 20% FCS, for 20–24 h. The bloodpurified trypomastigotes were added to the macrophagemonolayer at five parasites per cell and the preparationsincubated at 378C, 5% CO2 in DMEM 10% FCS. After 3 hthe monolayers were repeatedly washed with DMEM mediumto remove non firmly adhered or non penetrating trypo-mastigotes. Part of the preparations were immediately fixedin Bouin solution, washed with water and stained withGiemsa, whereas the remaining preparations were incubatedfor a further 24 h and then stained. The percentage of macro-phages infected was determined by examining 200 cellsat random (×1000). The mean number of intracellularparasites was determined in 100 infected cells.

Macrophage cultures

In order to assess the effect of bloodstream trypomastigotes,WT and BZR T. cruzi populations, on induction of NO,IL-12 and TNFa synthesis, mouse inflammatory macrophageswere resuspended in DMEM 5% FCS at 2×106 cells/ml,and 0·1 ml aliquots dispensed into wells of a 96-well tissueculture plate (Corning, NY, USA). The cells were allowed toadhere at 378C in 5% CO2. After 3 h incubation the cellswere washed once with DMEM at room temperature and0·1 ml of DMEM 10% FCS was added to each well. Differentconcentrations of bloodstream trypomastigotes of WT andBZR populations isolated, from mice treated and untreatedwere added to the macrophages in the presence or absence of100 U IFN-g/ml (Genzyme Corp., Cambridge MA, USA).The macrophages and parasites were incubated at 378Cin 5% CO2 in DMEM 10% FCS in a final volume of200ml/well. A positive control containing 100 ng/ml of

bacterial endotoxin (LPS-lipopolysaccharide) was intro-duced. Supernatants were collected after 12, 24 and 48 h ofculture for NO, IL-12 and TNFa measurements.

Splenocyte cultures

Mice infected with WT and BZR population were sacrificedat the peak of parasitemia 5, 24 and 48 h after treatment with500 mg/kg of BZ. Mice treated (n¼ 4) and non treated(n¼ 4), infected with each population had their spleensremoved. Suspensions of splenocytes were washed inDMEM and treated for 2 min with lysing buffer (9 vol. of0·16M NH4Cl and 1 vol. of 0·17M Tris-HCl, pH 7·5). Theerythrocyte-free cells were then washed three times andadjusted to 5×106 cells/ml in DMEM supplemented with10% FCS. The cell suspension was distributed (100ml/well)in a 96-well tissue culture plate and cocultured with DMEMalone, concanavalin A (ConA, 5mg/ml) or trypomastigotetissue culture antigen (10mg/ml) for 48 h at 378C in 5% CO2.The supernatants were subsequently collected for IFN-g andNO measurements. TheT. cruziantigen was obtained fromtrypomastigotes of WTT. cruzi, Y strain, maintained in L929cells (ATCC CCL-1). Trypomastigotes were suspended at108 cells/ml in a hypotonic buffer (10 mM NaCl, 1·5 mM

MgCl2, 1 mM dithiothreitol and 10 mM Tris-HCl, pH 7·4),freeze-thawed and centrifuged at 10 000g for 30 min. Thesupernatant was collected and protein concentration wasdetermined by the method of Bradford (Bradford 1976).

Determination of nitrite, IL-12, TNF a and IFN-g levelsin supernatants of macrophages and splenocytes

Nitrite concentration in the culture media was assayed byGriess reaction (Drapieret al. 1988). Plates were read at490 nm and NO2 concentration was determined with refer-ence to a standard curve using sodium nitrite in culturemedia. The minimal concentration of nitrite detectable was2 nmol/ml. IL-12 measurements was determined using a pairof MAbs against the p40 polypeptide of IL-12 (IL-12p40),which is tightly regulated in macrophages. Enzyme-linkedimmunosorbent assay (ELISA) was performed using 5mg/mlof antip40 MAb (C17·15·10) as the capture antibody andbiotinylated anti-IL-12 (C15·6.76), diluted 750-fold, as thedetecting antibody. The development was made with strepto-avidin-peroxidase conjugate. The plates were read at 405 nmand IL-12 (p40) concentration was calculated by referenceto a standard curve for murine rIL-12. TNFa was quantifiedin 48 h supernatants by ELISA using the Genzyme douset kit(Genzyme). IFN-g levels were determined by a two-sitesandwich ELISA using purified anti-Mu IFN-g (5 mg/ml)(Cytimmune, Lee Biomolecular Research Inc, San Diego,CA, USA) as the capture antibody and a rabbit anti-Mu

Volume 21, Number 10, October 1999 Effect of drug treatment onT. cruzi–macrophage interaction

q 1999 Blackwell Science Ltd,Parasite Immunology, 21, 535–544 537

IFN-g polyclonal serum (Immunex Corp., Seattle, WA,USA), diluted 1000-fold, as detecting antibody. The devel-opment was made with a rabbit anti-IgG whole moleculeperoxidase-conjugate. A standard curve was obtained withrMu IFN-g. The minimal concentration of each cytokinedetectable in the conditions of our assays was 1·5 ng/mlfor IL-12, 0·15 ng/ml for TNFa and 0·7 ng/ml for IFN-g.

Statistical analysis

Data were analysed by Student’st-test performed by usingINSTAT software (GraphPad, San Diego, CA, USA). AP-value of less than 0·05 was considered statistically significant.

RESULTS

Effect of BZ on the reduction of parasitemia

The parasitemia of mice infected with WT or BZR popula-tion or clones was evaluated at different times after a singlehigh dose of BZ (Figure 1). A reduction in parasitemia of99% was observed after 6 h of treatment for WT populationand clone treated with BZ as compared to untreated animals.In contrast, no reduction was detected for BZR populationand clone after treatment with BZ.

Molecular characterization of WT and BZR T. cruzipopulations and clones

The WT and BZRT. cruziparental populations and 10 clonesof either population used in this study were all classified asZymodeme Z2 for the six enzymes analysed (data notshown). RAPD has been used for analysis of genetic varia-tion and the identification of genetic markers (Steindelet al.1993). Thus, we used this technique to study the geneticvariability of susceptible and drug-resistantT. cruziclones.Complex and primer specific RAPD profiles were obtainedusing the three random primers (Figure 2a). The profilesproduced by the parental strains and clones were identicalindependent of BZ susceptibility, suggesting a high genomichomogeneity. Figure 2(b) shows the PCR amplificationproducts of TcPGP and HGPRT genes from WT and BZRpopulations and clones. The parental strains and clonespresented an identical monomorphic pattern, regardless ofthe drug susceptibility, reinforcing the genomic similarityfound by RAPD and isoenzyme analyses.

Effect of the host treatment with BZ on the phagocytosisof WT and BZR populations and clone by mouseinflammatory macrophages

The phagocytosis of WT population and clones isolatedfrom infected mice 3 h after treatment with BZ (500 mg/kg)

was significantly enhanced as compared to the same para-sites collected from untreated mice (Figure 3). The macro-phages clearly displayed a selective trypanocidal effect onWT population and clone as demonstrated by the dramaticdecline in the percentage of infected cells after 24 h incuba-tion. No trypanocidal effect was observed on BZR popula-tion and clone. These results were also confirmed by thedecrease in number of intracellular WT parasites observedat 3 h and 24 h after parasite phagocytosis. Approximatelyfive parasites/infected macrophage were found after 3 h andone parasite/infected macrophage at 24 h (data not shown).Regardless of being from a population or clone, the WTparasites obtained from treated mice were better phago-cytosed and destroyed by macrophages as compared to thesame parasite obtained from untreated animals. No effect ofBZ treatment was observed in terms of phagocytosis of theresistantT. cruzipopulation or clone. Observations at opticalmicroscopy showed alterations in shape and decrease inmotility in the WT but not in the BZR parasites isolatedfrom mice after treatment with BZ. However, no morpho-logical change was observed in organelles from WT para-sites obtained from treated animals, as evaluated by electronicmicroscopy (data not shown). As control, live or heat killed(2 min, 568C) WT parasites were incubated with inflam-matory macrophages and used to study phagocytosis andparasite destruction. No difference on phagocytosis of live



150

100

50

0

Par

asit

emia

(%

)

0

Time after treatment (h)

WT population not treatedWT population treatedBZR population not treatedBZR population treatedWT clone 11 not treatedWT clone 11 treatedBZR clone 27 not treatedBZR clone 27 treated

T. cruzi

2 4 6

Figure 1 Parasitemia changes after a single high-dose treatment withBZ (500 mg/kg, oral) in mice infected with the WT or BZRT. cruzipopulations and clones.

or heat-killed WT parasites isolated from untreated micewas observed. In contrast to live parasites, phagocytosis ofheat-killed trypomastigotes, recovered from treated mice,was very low. The heat-killed WT parasites, regardless ofBZ treatment, were more destroyed by macrophages thanthe live WT parasites (data not shown).

Induction of NO, IL-12 and TNF a production byT. cruzi-infected mouse macrophages primed withIFN-g

The induction of NO, IL-12 and TNFa synthesis by IFN-gprimed macrophages exposed to different numbers of

T. cruzi trypomastigotes was evaluated at different timespoststimulation. Thus, inflammatory macrophages wereexposed to five parasites/macrophage in presence of IFN-g

(100 U/ml) for 12, 24 and 48 h. The induction of the highestlevels of NO, IL-12 and TNFa in the tissue culture super-natants were reached after 48 h of parasite/macrophageinteraction (data not shown). Further, the ability of WT andBZR populations, isolated from infected mice untreated ortreated with BZ, to induce NO, IL-12 and TNFa productionwas evaluated after 48 h using different parasite/macrophageratios (Figure 4). The levels of NO, IL-12 and TNFa werenearly proportional to the parasite/macrophage ratio. Inter-estingly, BZR was a poor inducer of NO, IL-12 and TNFa,



Figure 2 (a) RAPD profiles of WT and resistant BZRT. cruziparental and five clones amplified with 3302, 3303 and 3307 primers. (b) PCRamplification products of TcPGP and HGPRT genes of WT and BZRT. cruziparental and two clones. The gels are 6% acrylamide silver staining.The numbers on the left indicate the size markers used (v ×174 digested withHaeIII).

regardless of previous treatment with BZ. More importantly,WT population isolated from mice 3 h after treatment withBZ induced the production of two times more NO (Figure 4a)and IL-12 (Figure 4b) and seven times more TNFa (Figure4c) than the same parasites isolated from untreated mice. BZdirectly added to the macrophage culture (1 ng to 1 mg/ml)did not induce NO, IL-12 and TNFa production by macro-phages primed with IFN-g (data not shown). As control, wecompare the ability of live vs. heat-killed WT parasites ininducing NO synthesis by macrophages. The heat-killingtreatment (2 min, 568C) of WT parasites resulted in NOreduction of three and four-fold, respectively, for untreatedand treated mice as compared to live parasites.

IFN-g and NO production in vivo by splenocytes frommice infected with WT and BZR population

In order to evaluate the capacity of susceptible and resistantparasite populations to stimulate the immune systemin vivo,

splenocytes from infected mice were isolated at peak para-sitemia, 5 h after treatment with 500 mg/kg of BZ. We firstmeasured the IFN-g levels, since this cytokine has beenreported to be inducedin vivoandin vitro by IL-12 producedby macrophages exposed toT. cruziproducts (Alibertet al.1996, Camargoet al. 1997). Figure 5 shows the ex-vivoIFN-g production by splenocytes cultured 48 h in the presenceof medium, ConA (5mg/ml) or T. cruziantigens (10mg/ml).



WT population not treatedWT population treatedBZR population not treatedBZR population treatedWT clone not treatedWT clone treatedBZR clone not treatedBZR clone treated

90

80

70

60

50

40

30

20

10

0

Per

cen

tag

e o

f in

fect

ed m

acro

ph

ages

3

Time after infection (h)

24

*

*T. cruzi

Figure 3 T. cruziphagocytosis and destruction by macrophages.Mouse inflammatory macrophages were infected with bloodtrypomastigotes from WT and BZR population and clones collectedfrom infected mice 3 h after treatment with BZ (500 mg/kg).Observations were made 3 and 24 h after macrophage infection byT. cruzi. The results are the mean and the standard deviation of threedifferent experiments in duplicate. *P<0·01 compared with the sameparasites collected from untreated mice.

****

100

80

60

40

20

0

Nit

rite

(n

mo

l/ml)

0 1 2 3 4 5

(a)

*

***

**

50

40

30

20

10

0

IL-1

2 (n

g/m

l)

0 1 2 3 4 5Parasite/macrophage ratio

(b)*

**

*

876543210

TN

F-α

(ng

/ml)

0 1 2 3 4 5

(c)*

*

*

WT treatedBZR not treatedBZR treated

*

Figure 4 Induction of nitrite (a), IL-12 (b) and TNFa (c) productionby T. cruzi-infected mouse inflammatory macrophages stimulated byIFN-g. The macrophages were infected with blood trypomastigotesfrom WT and BZR populations. The trypomastigotes were isolatedfrom infected mice 3 h after treatment with BZ (500 mg/kg) and nottreated. Inflammatory macrophages were exposed for 48 h at differentparasite/macrophage ratios in presence of 100 U/ml of IFN-g. Theresults are the mean and the standard deviation of three differentexperiments in duplicate. *P<0·01 compared with the same parasitescollected from untreated mice. **P<0·02 compared with BZR fromuntreated mice.

9080706050403020100

INF-

γ (n

g/m

l)

NM nottreated

NMtreated

WT nottreated

WTtreated

BZR not

treatedBZR

treated

StimulusCulture mediumConcanavalin AT. cruzi antigen * *

*

Figure 5 IFN-g production by splenocytes from normal mice andmice infected with WT or BZRT. cruzipopulations. The splenocyteswere isolated from infected mice at the peak parasitemia, 5 h aftertreatment with BZ (500 mg/kg). Controls were made isolatingsplenocytes from untreated infected and uninfected mice. Thesplenocytes were cultured for 48 h in the presence of medium alone,Con A (5mg/ml) or trypomastigote antigen (10mg/ml). Supernatantswere collected and IFN-g quantified. The results are expressed asmean and the standard deviation of three different experiments intriplicate. NM, normal mice; WT, wild-type and BZR, resistantT. cruzipopulations, respectively *P<0·01, compared withsplenocytes collected from mice infected with WT untreated and,BZR treated and untreated.

Regardless of the stimulus, the levels of IFN-g produced bysplenocytes from mice infected with WT population pre-viously treated with BZ were significantly enhanced. Incontrast, low levels of IFN-g were produced by splenocytesfrom mice infected with the BZR population, regardless oftreatment previous with BZ. Splenocytes from normal micetreated or untreated, also produced low levels of IFN-g

which were enhanced in the presence of Con A, as comparedwith culture medium alone. BZ (1 ng to 1 mg/ml) directlyadded to the splenocytes culture did not affect IFN-g andNO production (data not shown).

The NO synthesis is also dependent on monokine synth-esis (i.e. TNFa) by macrophages. Therefore we comparedthe production of NO by unstimulated splenocytes obtainedfrom untreated or treated infected mice. Figure 6 shows thein-vivo NO production by splenocytes harvested at 5, 24 and48 h post-treatment and cultured for 48 h in the presence ofmedium. The level of NO produced by splenocytes frommice infected with the WT population and treated with BZenhanced significantly at 24 and 48 h post-treatment, ascompared with splenocytes from untreated mice infectedwith the same parasites. On the other hand, the NO pro-duction by splenocytes from mice infected with the BZRpopulation, decreased progressively. Splenocytes fromuninfected mice treated or untreated, did not produce NO(data not shown).

DISCUSSION

One of our major interests is to understand the mechanism ofaction of resistance to in-vivo chemotherapy against the

intracellular protozoanT. cruzi. In our recent studies, 45T. cruzistrains susceptible and naturally resistant to BZ andnifurtimox were analysed for different molecular markers.The heterozygous profile, zymodeme B, contained exclu-sively susceptible strains, and occurred predominantly ingeographical areas where clinical treatment of Chagas’disease has been reported as more effective (Murtaet al.1998). In addition, these studies demonstrated that allT. cruzistrains naturally resistant to in-vivo treatment withnitroheterocyclic derivatives, presented a monomorphicprofile for a number of different molecular markers. Wehave selected in-vivo aT. cruzi population and severalclones, which are highly resistant to BZ treatment (Murta& Romanha 1998). In the present study, the molecularcharacterization of WT and BZR populations and clonesshowed that they were all genetically homogeneous usingthe markers employed. Because of their marker identity withthe parental strain, the selected resistant parasite clones arean excellent tool to identify parasite gene(s) involved inresistance as well as to study the in-vivo mechanism ofaction of nitroheterocyclic derivatives againstT. cruzi.

Little is known about the in-vivo mechanism of action ofBZ in T. cruzi. Nevertheless, studies suggest that theefficacy of drug treatment appears to be lower in immuno-suppressed animals. Toledoet al. (1991) demonstrated thatcyclophosphamide-induced immunosupression decreasescure rates in mice inoculated with differentT. cruzistrainsand submitted to specific treatment with BZ. In agreementwith these findings, a previous study (Michailowskyet al.1998) demonstrated that in-vivo neutralization of endo-genous IFN-g or IL-12 decreases the efficacy of BZ treat-ment during acute phase of experimental Chagas’ disease. Infact, it is well documented that treatment with nitrohetero-cyclic derivatives is more effective during the acute phase ofChagas’ disease, when a strong activation of the cellularcompartment of the immune system and release of high levelsof IFN-g, IL-12 and other pro-inflammatory cytokines isobserved (Brener & Gazzinelli 1997). Thus, a cooperativeeffect between nitroheterocyclic derivatives and the hostimmune system, during treatment againstT. cruzi, has beenproposed. In the present study, we compared the interactionbetween parasites and host macrophages using selected BZ-resistant and BZ-susceptible Y strain ofT. cruzi.

According to previous study using the Y strain ofT. cruzi(Lages-Silvaet al. 1990), our results demonstrate that in-vitro phagocytosis by murine macrophages is significantlyincreased when WTT. cruzipopulation or clones are obtainedfrom infected mice 3 h after BZ treatment. It may as well bethat the enhanced NO and cytokine synthesis is a directeffect of augmented phagocytosis of WT parasites obtainedafter in-vivo treatment with BZ. No effect of BZ treatmentwas observed on phagocytosis and cytokine synthesis, when



5

4

3

2

1

0

Incr

easi

ng

of

nit

rite

pro

du

ctio

n(t

reat

ed/u

ntr

eate

d)

5 Time after BZ treatment (h)

24 48

T. cruzi populationWTBZR

**

Figure 6 Increasing of nitrite production by splenocytes from miceinfected with WT and BZRT. cruzipopulations and treated with BZ.The splenocytes were isolated from untreated infected mice at peakparasitemia, as well as at 5, 24 and 48 h after treatment with BZ(500 mg/kg). Thereafter the splenocytes were cultured for 48 h andhad the nitrite quantified. The increasing of nitrite production isshown as the ratio of nitrite produced by splenocytes from mice witheitherT. cruzipopulations, treated/untreated. The results areexpressed as mean6 SD of two different experiments in triplicate.*P<0·001 WT compared with BZR population at the same time.The range of nitrite detection varied from 2 to 34·5 nmol/ml.

BZR population was used. Alternatively,in vivoexposure ofT. cruzi trypomastigotes WT parasites to BZ may result insome molecular and/or structural change(s) that lead toenhanced macrophage activation. Our preliminary resultsobtained by optical and electronic microscopy show thatin-vivo BZ treatment induce some change in motility andparasite shape. However, no morphological change wasobserved in organelles from WT parasites obtained fromtreated animals, as evaluated by electronic microscopy. Inaddition, earlier studies performed by Fonteset al. (1991)demonstrate that BZ-treatment of mice infected with suscept-ible and naturally resistantT. cruzistrains does not affect thecarbohydrate distribution in the membrane of the parasite.Current studies are being performed in our laboratory, inorder to define the molecular and structural surface changesobserved on the BZ-susceptibleT. cruziafter in-vivo treat-ment with BZ.

In addition to enhanced phagocytosis, the results presentedhere show thatin vitro, WT parasites isolated from micetreated with BZ induced an enhanced production of NO andthe cytokines IL-12 and TNFa, by mouse macrophageprimed with IFN-g. The in-vivo importance of this phenom-enon is shown by ex-vivo experiments using splenocytesfrom infected animals treated or not treated with BZ. Theseexperiments show that splenocytes obtained from miceinfected with WT parasites and treated with BZ, producedhigher levels of IFN-g, than splenocytes from non treatedinfected mice. Moreover, we observed a dramatic increaseof NO synthesis by splenocytes from animals infected withthe WT parasites treated with BZ and removed at 24 h and48 h after treatment, as compared with splenocytes frominfected but untreated animals. In contrast, the BZR popula-tion induced low levels of NO, IL-12, TNFa and IFN-g,regardless of previous treatment with BZ.

Although the immune response, can control parasitereplication during acute phase of infection, alone it is notsufficient to eliminate theT. cruzi. Nevertheless, our studiessuggest that in combination with BZ, the immune systemcan eliminate the parasite from the host tissues. It is possiblethat someT. cruzistrains can more effectively stimulate theimmune response than others (Mulleret al. 1986). In fact,studies performed in our laboratory suggest that naturallyresistant strains (i.e. Colombiana and VL-10) ofT. cruziarepoor inducers of IL-12 synthesis by macrophages bothinvitro and in vivo (Michailowsky et al. 1998). These resultshave now been confirmed by using BZ-susceptible andresistantT. cruzipopulations derived from the same strain.Both parasite populations are genetically homogeneousbased in the molecular markers analysed, but immunologi-cally different, since WT parasites from untreated miceinduced significantly higher levels of NO and IL-12 inmacrophages, as compared to BZR parasites. These data

suggest that the WT population may have more molecules,which are able to induce the production of IL-12 and NO bymacrophages. In addition, the BZ treatment may uncoverthe remaining stimulator molecules, further increasing theproduction of IL-12 and NO, favouring BZ activity againstT. cruzi. Alternatively, as shown by our in-vitro experi-ments, it is possible that the treatment with BZ rendersdrug-susceptible parasites more available to macrophagephagocytosis, enhancing the interaction and stimulation ofmacrophages by the parasite products. This stimulation ofmacrophage will then favour the immune attack to theremaining circulating parasites.

The interaction, of immune system and chemotherapyagainst parasites has been previously reported in differentstudies (Target 1985, Berger & Fairlamb 1992). An antibodyresponse to surface antigens is necessary to remove Africantrypanosomes from the blood of infected mice treated withDL-a-difluoromethylornithine (Degeeet al. 1983). The effectof praziquantel onSchistosoma mansonialso depends on thesynergistic interaction between the drug and the host immuneresponse (Brindley & Sher 1987). More recently, differentstudies have shown that IL-12 potentiates the action ofchemotherapy against different protozoa (i.e.L. major,Plasmodiumsp. andT. cruzi) (Naborset al. 1995, Stevensonet al. 1995, Michailowskyet al. 1998); fungi (Cryptococcusspp. andHistoplasma capsulatum) (Clemonset al. 1994,Zhouet al. 1997) and bacteria (Mycobacterium tuberculosis)(Flynn et al. 1995).

Finally, comparative studies of drug susceptibility onT. cruzi, demonstrate no correlation between parasite drugsusceptibilityin vitro and in vivo (Scoth & Mathews 1987,Neal & Van Bueren 1988; Ribeiro-Rodrigueset al. 1995). Infact, in-vitro susceptibility of bloodstream trypomastigotesin macrophage cultures do not distinguish theT. cruzistrainswhich are susceptible or resistant to in-vivo treatment withnifurtimox or BZ (Neal & Van Bueren 1988). These studiessuggest the involvement of the immune system in the efficacyof treatment ofT. cruziinfections. Taken together, our resultssuggest the existence of a cooperation between BZ treatmentand the immune response againstT. cruzi, which ensure theefficacy of drug therapy. Interestingly, this cooperative effectis observed on the susceptible but not on the drug-resistantT. cruzipopulation and may explain, in part, a mechanism bywhich T. cruziparasites may resist to chemotherapyin vivo.

ACKNOWLEDGEMENTS

This work received financial support from CNPq (ConselhoNacional de Pesquisas), PRONEX (Programa de Apoio aNucleos de Exceleˆncia) no. 2704, PAPES A–FIOCRUZ(Programa de Apoio a` Pesquisa Estrate´gica) and FAPEMIG(Fundac¸ao de Amparo a` Pesquisa do Estado de Minas



Gerais). We thank Dr Philip Loverde for the Englishcorrection and suggestions.

REFERENCES

Alibert J.C.A., Cardoso M.A.G., Martins G.A.et al. (1996) IL-12mediates resistance toTrypanosoma cruziin mice and is producedby murine macrophages in response to live trypomastigotes.Infectionand Immunity, 64, 1961–1967

Berger B.J. & Fairlamb A.H. (1992) Interactions between immunity andchemotherapy in the treatment of the trypanosomiases and leish-maniases.Parasitology, 105,871–878

Bradford M.M. (1976) A rapid and sensitive method for the quantitationof microgram quantities of protein utilizing the principle of protein-dye binding.Annals of Biochemistry, 72, 248–254

Brener Z. (1962) Therapeutic activity and criterion of cure on miceexperimentally infected withTrypanosoma cruzi. Revista do Institutode Medicina Tropical de Saõ Paulo, 4, 389–396

Brener Z. & Gazzinelli R.T. (1997) Immunological control ofTrypano-soma cruziinfection and pathogenesis of Chagas disease.Inter-national Archives of Allergy and Immunology, 114,103–110

Brindley P.J. & Sher A. (1987) The chemotherapeutic effect of prazi-quantel againstSchistosoma mansoniis dependent on host antibodyresponse.Journal of Immunology, 139,215–220

Camargo E.P. (1964) Growth and differentiation inTrypanosoma cruzi.I. Origin of metacyclic trypanosome in liquid media.Revista doInstituto de Medicina Tropical de Saõ Paulo, 6, 93–100

Camargo M.M., Almeida I.C., Pereira M.E.t al. (1997) Glycosylpho-sphatidylinositol-anchored mucin-like glycoproteins isolated fromTrypanosoma cruzitrypomastigotes initiate the synthesis of proinflam-matory cytokines by macrophages.Journal of Immunology, 158,5890–5901

Cardillo F., Voltarelli J.C., Reed S.G. & Silva J.S. (1996) Regulation ofTrypanosma cruziinfection in mice by gamma interferon andinterleukin-10: the role of NK cells.Infection and Immunity, 64,128–134

Carneiro M., Chiari E., Gonçalves A.M.et al. (1990) Changes in theisoenzyme and kinetoplast DNA patterns ofTrypanosoma cruzistrains induced by maintenance in mice.Acta Tropica, 47, 35–45

Clemons K.V., Brummer E. & Stevens D.A. (1994) Cytokine treat-ment of central nervous system infection: efficacy of IL-12 aloneand synergy with conventional antifugal therapy in experimentalcryptococcoses.Antimicrobial Agentsand Chemotherapy,38,460–464

Degee A.L.W., McCann P. & Mansfield J.M. (1983) Role of antibody inthe elimination of trypanosomes after DL-a-difluoromethylornithinechemotherapy.Journal of Parasitology, 69, 818–822

Drapier J.C., Wietzerbin J. & Hibbs J.B. Jr (1988) Interferon-g andtumor necrosis factor induce theL-arginine-dependent cytotoxiceffector mechanism in murine macrophages.European Journal ofImmunology, 18, 1587

Filardi L.S. & Brener Z. (1984) A rapid method for testing in vivo thesusceptibility of different strains ofTrypanosoma cruzito activechemotherapeutic agents.Memorias do Instituto Oswaldo Cruz, 79,221–225

Filardi L.S. & Brener Z. (1987) Susceptibility and natural resistance ofTrypanosoma cruzistrains to drugs used clinically in Chagas’disease.Transactions of the Royal Society of Tropical Medicineand Hygiene, 81, 755–759

Flynn J.L., Goldstein M.M., Triebold K.J.et al. (1995) IL-12 increasesresistance of Balb/c mice toMycobacterium tuberculosisinfection.Journal of Immunology, 155,2515–2524

Fontes G., Romanha A.J., Pereira M.E. & Brener Z. (1991) Lectinreceptors distribution in the surface membrane ofTrypanosoma cruziblood forms collected from mice submitted to specific treatment.Memorias do Instituto Oswaldo Cruz, 86, 297–299

Gazzinelli R.T., Hieny S., Wynn T.A., Wolf S. & Sher A. (1993)Interleukin-12 is required for the T-lymphocyte-independent induc-tion of interferon-gamma by an intracellular parasite and inducesresistance in T-cell deficient hosts.Proceedings of the NationalAcademy of Sciences USA, 90, 6115–6119

Gazzinelli R.T., Oswald P., Hieny S., James S.L. & Sher A. (1992) Themicrobicidal activity of interferon-g-treated macrophages againstTrypanosoma cruzi involves anL-arginine-dependent, nitrogenoxide-mechanism inhibitable by interleukin-10 and transforminggrowth factor-b. European Journal of Immunology, 22, 2501–2506

Hunter C.A., Slifer T. & Araujo F. (1996) Interleukin-12 mediatedresistance toTrypanosoma cruziis dependent on tumor necrosisfactor alpha and gamma interferon.Infection and Immunity, 64,2381–2386

Khaw M. & Panosian C.B. (1995) Human antiprotozoal therapy:past, present, and future.Clinical Microbiology Review of, 8, 427–439

Lages-Silva E., Filardi L. & Brener Z. (1990) Effect of the host specifictreatment in the phagocytosis ofTrypanosoma cruziblood forms bymouse peritoneal macrophages.Memorias do Instituto OswaldoCruz, 85, 401–405

Michailowsky V., Murta S.M.F., Carvalho-Oliveira L.et al. (1998)Interleukin-12 enhances in vivo parasiticidal effect of benznidazoleduring acute experimental infection with a naturally drug-resistantstrain of Trypanosoma cruzi. Antimicrobial Agents and Chemo-therapy, 42, 2549–2556

Muller L.A., Anasco N. & Gonzales Cappa S.M. (1986)Trypanosomacruzi: isolate dependence in the induction of lytic antibodies in themouse.Experimental Parasitology, 61, 284–293

Murta S.M.F., Gazzinelli R.T., Brener Z. & Romanha A.J. (1998)Molecular characterization of susceptible and naturally resistant strainsof Trypanosoma cruzito benznidazole and nifurtimox.Molecularand Biochemical Parasitology, 93 (2), 203–214

Murta S.M.F. & Romanha A.J. (1998) In vivo selection of a populationof Trypanosoma cruziand clones resistant to benznidazole.Para-sitology, 116,165–171

Nabors G.S., Alonso L.C.C., Farrell J.P. & Scott P. (1995) A switchfrom a Th2 to Th1 type response and cure of establishedLeishmaniamajor infection in mice is induced by combined therapy withinterleukin 12 and pentostam.Proceedings of the National Academyof Sciences, USA, 92, 3142–3146

Neal R.A. & Van Bueren J. (1988) Comparative studies of drugsusceptibility of five strains ofTrypanosoma cruziin vivo and invitro. Transactions of the Royal Society of Tropical Medicine andHygiene, 82, 709–714

Pereira da Silva L.H.P. & Nussenzweig V. (1953) Sobre uma cepa deTrypanosoma cruzialtamente virulenta para camundongo branco.Folia Clinica et Biologica, 20, 191–207

Reed S.G. (1988) In vivo administration of recombinant IFN-gammainduces macrophage activation, and prevents acute disease, immunesuppression, and death in experimentalTrypanosoma cruziinfec-tions.Journal of Immunology, 140,4342–4347

Ribeiro-Rodrigues R., Dos Santos W., Oliveira A.B.et al. (1995)Growth inhibitory effect of nafthofuran and nafthofuranquinonederivatives onTrypanosoma cruziepimastigotes.Bioorganic andMedicinal Chemistry Letters, 5, 1509–1512

Scoth V.R. & Mathews T.R. (1987) The efficacy of an n-substitutedimidazole, RS-49676, against aTrypanosoma cruziinfection in



mice. American Journal of Tropical Medicine and Hygiene, 37,308–313

Silva J.S., Morrissey P.J., Grabstein K.H.et al. (1992) Interleukin10 and interferon gamma regulation of experimentalTrypanosomacruzi infection.Journal of Experimental Medicine, 175,169–174

Silva J.S., Vespa G.N.R., Cardoso M.A.G., Alibert J.C.S. & Cunha F.Q.(1995) Tumor necrosis factor alpha mediates resistance toTrypano-soma cruziinfection in mice by inducing nitric oxide production ininfected gamma interferon-activated macrophages.Infection andImmunity, 63 (12), 4862–4867

Steindel M., Dias Neto E., Menezes C.L.P., Romanha A.J. & SimpsonA.J.G. (1993) Random amplified polymorphic DNA analysis ofTrypanosoma cruzistrains.Molecular and Biochemical Parasitology,60, 71–80

Stevenson M.M., Tam M.F., Wolf S.F. & Sher A. (1995) IL-12-inducedprotection against blood-stagePlasmodium chabaudirequires IFNgand TNFa and occurs via nitric oxide-dependent mechanism.Journal of Immunology, 156,263–268

Target G.A.T. (1985) Chemotherapy and the immune response inparasitic infections.Parasitology, 90, 661–673

Toledo M.J.O., Machado G.B.N., Pereira M.E.S. & Brener Z. (1991)Results of treatment in mice immunosuppressed inoculated withdifferentTrypanosoma cruzistrains.Memorias do Instituto OswaldoCruz, 86, 237

Torrico F., Heremans H., Rivera M.T.et al. (1991) Endogenous IFN-gamma is required for resistance to acuteTrypanosoma cruziinfec-tion in mice.Journal of Immunology, 146,3626–3632

Vespa G.N.R., Cunha F.Q. & Silva J.S. (1994) Nitric oxide isinvolved in the control ofTrypanosoma cruziinduced parasitemiaand directly kills parasite in vitro.Infection and Immunity, 62,5177–5182

World Health Organization (1994) Chagas Disease. WHO TechnicalReport Series, TDR News, p. 125–134

Zhou P., Sieve M.C., Tewari R.P. & Seder R.A. (1997) Interleukin-12modulates the protective immune response in SCID mice infectedwith Histoplasma capsulatum. Infection and Immunity, 65, 1–7



in-vivo treatment with benznidazole enhances phagocytosis, parasite destruction and cytokine release...

Documents