role of superoxide and reactive nitrogen intermediates in rhodnius prolixus (reduviidae)/...
TRANSCRIPT
Experimental Parasitology 98, 44–57 (2001)
doi:10.1006/expr.2001.4615, available online at http://www.idealibrary.com on
Role of Superoxide and Reactive Nitrogen Intermediates in Rhodnius prolixus(Reduviidae)/Trypanosoma rangeli Interactions
M. M. A. Whitten,* C. B. Mello,† S. A. O. Gomes,‡ Y. Nigam,§ P. Azambuja,‡ E. S. Garcia,‡and N. A. Ratcliffe*,1
*Biomedical and Physiological Research Group, School of Biological Sciences, University of Wales Swansea,Genchem104ark,
Index Descriptors and Abbreviations: Rhodnius prolixus; Trypano-soma rangeli; H14 and Choachi strains; hemolymph; NEM, N-ethylma-leimide; ?NO, nitric oxide; NADPH oxidase; NBT, nitroblue tetra-
Singleton Park, Swansea, SA2 8PP, United Kingdom; †Department ofFluminense, Niteroi, CEP 24.001-970, RJ, Brasil; ‡Department of BioFundacao Oswaldo Cruz, Avenida Brasil 4365, Rio de Janeiro, CEP 2§School of Health Sciences, University of Wales Swansea, Singleton P
Whitten, M. M. A., Mello, C. B., Gomes S. A. O., Nigam, Y.,Azambuja, P., Garcia, E. S., and Ratcliffe, N. A. 2001. Role of superox-ide and reactive nitrogen intermediates in Rhodnius prolixus(reduviidae)/Trypanosoma rangeli interactions. Experimental Parasi-tology 98, 44–57. This study compares aspects of the superoxide, nitricoxide and prophenoloxidase pathways in Rhodnius prolixus hemo-lymph, measured in parallel, in response to Trypanosoma rangeli inocu-lation. Responses to two strains of T. rangeli, and two developmentalforms, were studied, and the results obtained were correlated with theability of the parasites to survive, multiply, and complete their lifecycles in the hemolymph of the host. T. rangeli H14 strain parasites,which fail to complete their life cycle in Rhodnius by invading thesalivary glands, stimulated high levels of superoxide and prophenoloxi-dase activity, which peaked 24 h after inoculation. Simultaneously, theconcentration of hemolymph nitrites and nitrates increased, indicativeof nitric oxide activity, but parasite numbers remained low. T. rangeliChoachi strain parasite inoculation also stimulated superoxide andprophenoloxidase activity, which, though significantly lower than theequivalent responses to the H14 strain, also peaked at 24 h. However,nitrate and nitrite levels in Choachi strain-inoculated hemolymph re-mained low, and this parasite strain multiplied rapidly, especially fol-
lowing peak superoxide activity, and eventually invaded the salivaryglands for transmission to a vertebrate host. In both strains, short formepimastigotes stimulated greater superoxide and prophenoloxidase re-sponses than long form epimastigotes. Injection of the NADPH oxidaseinhibitor N-ethylmaleimide or the inducible nitric oxide synthase inhibi-tor S-methyl isothiourea sulfate caused significantly higher insect mor-talities in groups of R. prolixus inoculated with either parasite strain1To whom correspondence should be addressed. Fax: 01792 295447.E-mail: [email protected].
44
eral Biology, Universidade Federalistry and Molecular Biology,
5, RJ, Brasil; andSwansea, SA2 8PP, United Kingdom
compared with those of uninfected control insects. This indicates thatboth NADPH oxidase and nitric oxide synthase activity may be in-volved in the immune response of R. prolixus to infection by T. rangeli.Finally, Western blotting of R. prolixus hemocyte lysates revealed thepresence of a protein immunologically related to the human NADPHoxidase complex, the initiator enzyme of the respiratory burst.q 2001 Academic Press
zolium; NOS, nitric oxide synthase; O22 , superoxide; PpO,
prophenoloxidase; PTU, phenylthiourea; RNI, reactive nitrogen inter-mediate; ROS, reactive oxygen species; SMT, S-methyl isothioureasulfate; SOD, superoxide dismutase.
INTRODUCTION
The triatomine bug Rhodnius prolixus is an importantvector of Trypanosoma cruzi, the etiologic agent of Chagas’disease. In addition, it is also a host for Trypanosoma rangeli,
which is apparently nonpathogenic to humans but which canbe pathogenic to triatomines (Watkins 1971). Unlike T. cruzi,which develops in the gut of its insect vector and rarelyinvades the hemocoel (reviewed by Zeledon 1987; Mello etal. 1995), T. rangeli also develops in the gut, but commonlyinvades the hemolymph (Watkins 1971; Hecker et al. 1990).0014-4894/01 $35.00Copyright q 2001 by Academic Press
All rights of reproduction in any form reserved.
S
Anopheles gambiae, which is transcriptionally activated byboth bacteria and Plasmodium parasites, and it is also partic-ularly active in the cells of the midgut.
FREE RADICALS IN TRIATOMINE—TRYPANOSOME INTERACTION
Many aspects of the humoral and cellular defenses havebeen studied in trypanosome–triatomine interactions, in-cluding lysozyme and trypanolytic activity (Mello et al.1995), prophenoloxidase activation (Mello et al. 1995;Gomes et al. 1999), phagocytosis and nodule formation(Takle 1988; Mello et al. 1995), and agglutination (Pereiraet al. 1981; Mello et al. 1995; Ratcliffe et al. 1996). Oneaspect of triatomine immunity to trypanosomes, and of insectvector immunity in general, that has been neglected is therole of pathways involving free radicals such as superoxide(O2
2 ) and nitric oxide (?NO).In vertebrate phagocytic cells, O2
2 can be derived by theone-electron reduction of molecular oxygen by the so-calledrespiratory burst enzyme, NADPH oxidase. This reactionhas been extensively studied in vertebrates. Toxic reactiveoxygen species (ROS) derived from O2
2 provide a vital rolein antimicrobial (Babior et al. 1973; for reviews, see Gabigand Babior 1981; Wientjes and Segal 1995) and antiparasiticdefense (Nathan et al. 1979; reviewed by Hughes 1988).The ROS exert their oxidizing effects on target membranes,proteins, and DNA (reviewed by Hughes 1988; Wientjesand Segal 1995). T. cruzi infections are known to stimulatethe release of ROS from many vertebrate phagocytic cells,including human neutrophils (Docampo et al. 1983). In addi-tion, the chemotherapeutic effects of many antiparasiticdrugs, such as b-lapachone, are also attributed to free radicalactivity (Brisby 1990). Evidence for O2
2 and ROS activityin insect immunity is growing (Sun and Faye 1995; Cox-Foster et al. 1998; Fenimore and Cox-Foster 1998; Kobay-ashi 1998; Whitten and Ratcliffe 1999) and includes thedetection of ROS generated during melanization (prophenol-oxidase, PpO, activity) (Nappi et al. 1995; Fenimore andCox-Foster 1998; Nappi and Vass 1998b; Slepneva et al.1999). However, to date, no studies have addressed the roleof this free radical in the antiparasitic defenses of an insectvector of disease.
Another important free radical is nitric oxide, which isan ubiquitous molecule with a vast array of biological roles.Nitric oxide is now recognized as a crucial effector moleculein vertebrate immunology (reviewed by Fang 1997). In thiscontext, ?NO is generated by an inducible isoform of nitricoxide synthase (NOS) and reacts readily with other free
radicals, such as O22 , to generate peroxynitrite and other so-called reactive nitrogen intermediates (RNIs). Some of theseRNIs have potent antibacterial and antiparasitic properties(Zhu et al. 1992) and are known to be involved in responsesto T. cruzi infections in mammals (Vespa et al. 1994; Rotten-berg et al. 1996).
There is much evidence for ?NO activity in invertebrates(reviewed by Martınez 1995), but few studies are related to
45
immune function, and in R. prolixus, only a salivary NOSwith a vasodilatory role has been characterized (Yuda et al.1996). However, there are some reports of ?NO activity ininvertebrate immunity (Conte and Ottaviani 1995; Franchiniet al. 1995; Weiske and Wiesner 1999) and speculation thatinsect-derived ?NO and RNIs associate with ROS as part ofan immune response (Nappi and Vass 1998b). Luckhart etal. (1998) and Luckhart and Rosenberg (1999) have identi-fied and partially characterized an immune-reactive NOS inan insect vector of disease. The NOS they described in themidgut of the mosquito Anopheles stephensi is inducibleby, and able to limit, Plasmodium infections. In addition,Dimopoulos et al. (1998) have identified an NOS gene in
In order to elucidate the role, if any, of O22 and ?NO in
insect vector immunity, this paper compares the differentialactivation of the pathways leading to the formation of thesemolecules in R. prolixus hemolymph, in response to inocula-tion with two strains and developmental forms of T. rangeli.
MATERIALS AND METHODS
Chemicals All chemicals were obtained from Sigma ChemicalCompany (Poole, Dorset, UK) unless otherwise stated. The buffer usedthroughout was 30 mM Mops buffer [3-(N-morpholino)propanesulfonicacid, 380 mOsm, pH 7.8].
Insects. Fourth and fifth instar nymphs of R. prolixus were usedthroughout the experiments. The insects were reared at 268 6 28C,80% relative humidity, and a 12:12 h light–dark cycle, in plastic potscontaining filter paper and chicken feathers. Defibrinated horse blood(TCS Microbiology, Buckingham, UK) was fed to the insects fort-nightly using a membrane feeding system (Azambuja and Garcia 1997).Additionally, the diet was supplemented every 6 to 8 weeks with humancitrated blood from healthy volunteers.
T. rangeli and injections. T. rangeli strain H14 (supplied by Dr.C. B. Mello, FIOCRUZ, Brazil) was maintained at 268–288C in brain–heart infusion broth (BHI, Becton–Dickinson Microbiology Systems,Becton–Dickinson Europe, France), supplemented with 20% heat-inac-tivated fetal calf serum (Garcia and Azambuja 1997). T. rangeli strain
Choachi (supplied by Professor G. A. Schaub, Ruhr University, Bo-chum, Germany) was maintained at 268–288C in liver infusion tryptosemedium (LIT, Becton–Dickinson Microbiology Systems) supple-mented with 20% heat-inactivated fetal calf serum (Garcia and Azam-buja 1997).Short form H14 and Choachi epimastigotes were harvested duringlate log phase growth, typically 4 to 7 days after subculture, whilelong form epimastigotes were harvested during stationary phase growth,12 to 30 days after subculture for H14 strain T. rangeli and 14 to 30days for the Choachi strain. Trypanosomes were centrifuged at 800g,
46
10 min, at 268C, washed three times, and resuspended to a final concen-tration of 5 3 106 ml21 in sterile 30 mM Mops buffer. A 50-ml Hamiltonglass syringe with a short 30-gauge needle (Hamilton, Switzerland) wasfitted to a 1-ml microinjector (Hamilton, Switzerland), and hemolymphinfections were initiated by injection of 2-ml trypanosome suspensioninto the thorax by puncturing the articulation of a prothoracic coxa.Control insects were injected with 2 ml Mops buffer. Insects wereinjected 3 to 4 days after feeding.
Hemolymph collection and hemocyte/trypanosome counts. Insectswere lightly anesthetized on ice before bleeding. Hemolymph wascollected from 8 to 10 insects per treatment group, by cutting offmetathoracic legs and gently pressing the abdomen. The exuded hemo-lymph was collected in sterile 10-ml Pyrex microsampling pipettes(Corning) and immediately pooled into ice-cold, sterile 0.5-ml Eppen-dorf tubes. For all the assays except those testing PpO activity, thepooled hemolymph was immediately diluted 1:2 with a solution con-taining 1 part anticoagulant (AC):4 parts Mops buffer saturated withphenylthiourea (PTU) to prevent melanization. Anticoagulant com-prised 10 mM EDTA (free acid), 100 mM glucose, 79 mM sodiumchloride, 30 mM trisodium citrate, and 26 mM citric acid monohydrate,pH 4.6 (Leonard et al. 1985). For testing PpO activity, 60 ml ofhemolymph was expelled into tubes containing an equal volume of1:1 AC:Mops buffer (without PTU) and mixed gently. To conduct totalhemocyte and differential trypanosome counts, hemolymph was diluted10-fold in the latter AC/Mops/PTU solution and counted in a hemocy-tometer under a light microscope.
Superoxide assay. The O22 microplate assay was modified from
the method of Pick and Mizel (1981). The reduction of NBT by O22
generated insoluble blue formazan, and the absorbance was measuredat 550 nm. In order to assess the kinetics of O2
2 generation in responseto trypanosome or buffer injection, hemolymph was taken at 6, 12,24, 48, or 72 h after injection. Twenty-five microliters of fresh wholehemolymph collected from 10 T. rangeli- or control (buffer)-injectedinsects was added to 20 ml of 3 mg ml21 nitroblue tetrazolium (NBT)plus 30 ml Mops buffer in triplicate wells of a flat-bottomed 96-wellmicrotiter plate (Nunclon, Roskilde, Denmark) and incubated in a dark,humid chamber at 298C. Formazan absorbance was measured at 550nm at increasing time intervals (0, 10, 20, 30, 45, 60, and 120 min),using an Anthos Labtec HT II ELISA plate reader, but for simplicityonly values for 120 min are reported.
In replicates of the experimental wells, 5 ml of Mops (of the total30 ml) was replaced with 5 ml (375 U) of superoxide dismutase (CuZnSOD from bovine erythrocytes, EC 1.15.1.1), to demonstrate the pro-portion of NBT reduction attributable to O2
2 , for which SOD is aspecific antioxidant. In the blanks, 25 ml of Mops 1 PTU/AC replacedhemolymph. The positive control replaced hemolymph with UV-irradi-ated riboflavin (vitamin B2, 0.1 mg ml21, Shandon Southern ProductsLtd., Cheshire, UK), which autooxidizes to generate O2
2 (Segura-Aguilar 1993). The negative control replicated the positive control,except for the inclusion of 5 ml (375 U) of SOD instead of Mops buffer.
In addition, the in vitro hemolymph O22 responses to each strain and
form of T. rangeli (H14/Choachi, short/long form epimastigotes) weremeasured in hemolymph extracted from 10 naıve (not injected) insectsper treatment group. The contents of the experimental wells wereidentical to those described previously, except that 25 ml of T. rangeli(1 3 106 ml21) was added as an elicitor in the place of Mops buffer.As a further control, NBT reduction was also measured in naıve insecthemolymph in the absence of T. rangeli (i.e., buffer was used instead).
Hemocyte monolayers. Monolayers of hemocytes from R. prolixuswere made to visualize intracellular formazan staining. Hemolymph
WHITTEN ET AL.
was diluted 1:1 with Mops 1 PTU/AC and 25 ml was placed on sterilePTFE-coated multispot microscope slides (Hendley Ltd., Essex, UK),together with 20 ml of 3 mg ml21 NBT, plus or minus 20 ml of theelicitor Escherichia coli lipopolysaccharide (LPS; serotype 026:B6,Sigma, 1 mg ml21 dissolved in buffer). The slides were incubated ina dark and humid chamber at 298C for 30 min and gently washed byimmersion in a Coplin jar containing 30 mM Mops. The monolayerswere fixed in 4% formaldehyde in Mops buffer for 30 min and carefullywashed in Mops, and coverslips attached with Kaiser’s glycerin jellybefore sealing with nail varnish. Blue formazan deposits were observedunder bright-field optics.
PpO activating assay. Pooled hemolymph was collected on icefrom 10 uninjected or trypanosome- or buffer-injected insects, as de-tailed above. For each treatment group, two 60-ml aliquots of pooledhemolymph were immediately made. One aliquot was diluted 1:1 withMops, and the other aliquot was diluted 1:1 with Mops saturated withPTU. Thirty microliters each of PTU- and non-PTU-treated hemolymphwere incubated in triplicate flat-bottomed microtiter plate wells for 20min, in a humid chamber at 298C. Following addition of 10 ml of 3mg ml21 laminarin (b-1,3-glucan from Laminaria digitata, Sigma,dissolved in Mops buffer), the hemolymph was incubated for 30 minbefore adding 10 ml of saturated L-DOPA solution dissolved in deion-ized water and incubating for a further 2 h (optimal incubation period,Gregorio and Ratcliffe 1991). The absorbance of the melanin formedwas measured at 492 nm. In the blanks, buffered AC replaced hemo-lymph. Prophenoloxidase activity was expressed as the mean PTU-inhibitable change in OD/mg of hemolymph protein. Assays were runat 6, 12, 24, 48, or 72 h after injection, in order to assess the kineticsof PpO activity.
Electrophoresis and Western blotting for NADPH oxidase determina-tion in hemocytes. Western blotting was performed to determine ifR. prolixus hemocytes contained a protein with structural similaritiesto the mammalian NADPH oxidase. To stimulate activation of theputative NADPH oxidase complex, 25 R. prolixus were injected with2 ml of E. coli LPS (dissolved in insect saline; 18 g L21 D-glucose,12.2 g L21 potassium chloride, 0.36 g L21 sodium bicarbonate, 380mOsm, pH 7.8, adapted with modifications from Wiesner 1991), usingthe same injection method as for trypanosome inoculation. After 24h, whole hemolymph was collected from 25 injected insects and pooledinto a sterile 0.5-ml Eppendorf tube on ice, containing 1:4 AC:Mopsbuffer saturated with PTU. The hemolymph was centrifuged at 600g,48C, for 10 min, to pellet the hemocytes, which were then resuspendedand washed with AC. The plasma supernatant was retained on ice.Hemocytes were lysed by resuspension in sterile water and sonicatedon ice using 12 3 5 s bursts in a Soniprep 150 sonicator (FisherScientific, Manchester, UK). The plasma and crude lysates were finallyfreeze-dried (Modulyo freeze dryer, BOC Edwards, West Sussex, UK)and stored at 2208C in sealed tubes flushed with nitrogen gas.
To determine the presence of NADPH oxidase p67phox protein,freeze-dried crude hemocyte lysates were dissolved in concentratedreducing sample buffer (Laemmli 1970) and proteins were separated
on 12% SDS polyacrylamide gels which had been preelectrophoresedto remove residual peroxides, using a Bio-Rad minigel system (MiniProtean II Slab Cell, Bio-Rad Laboratories Ltd., Hertfordshire, UK).Human SKN cell lysate (derived from a human neuroblastoma withepithelial-like morphology, Transduction Laboratories, Kentucky,U.S.A.) was used as a positive control.Proteins were transferred to nitrocellulose membrane (Hybond CSuper, 0.45 mm, Amersham Life Science, Bucks, UK), by electroblot-ting at 20 V overnight, using a Bio-Rad Trans-Blot cell. Following
S
NBT and PpO assay data were analyzed using repeated measuresANOVA with Tukey’s multiple comparison posttest (according to Mo-tulsky 1995). The mortality of insects in the in vivo NOS inhibition
FREE RADICALS IN TRIATOMINE—TRYPANOSOME INTERACTION
blocking with 2% skim milk and 1% goat serum, the membranes wereincubated for 1 h at room temperature with 1:1000 rabbit polyclonalanti-p67phox primary antibody (Transduction Laboratories), raisedagainst a 17.3-kDa protein fragment corresponding to amino acids317–469 of human p67phox. After washing with 50 mM Tris-bufferedsaline (pH 7.4, 370 mOsm) containing 0.1% Tween, blots were thenincubated with 1:5000 alkaline phosphatase-conjugated goat anti-rabbitIgG (Sigma) for a further 1 h. Following washing, blots were developedwith 5-bromo-4-chloro-3-indoyl phosphate/nitroblue tetrazolium(BCIP/NBT) liquid substrate (Sigma) for 10–15 min. As a negativecontrol, membranes were incubated with 1:1000 rabbit serum (Sigma),instead of primary antibody, and then treated as above.
Nitrite and nitrate determinations. Pooled hemolymph samples(80 ml) from trypanosome- or insect saline-injected insects, preparedas above, were sonicated and tested for total nitrate and nitrite content,which can be indicative of RNI metabolism (reviewed by Moncada etal. 1991). Eighty-microliter samples were processed using a Griessassay kit (Cayman Chemical Co. MI, U.S.A.), involving enzymaticreduction of nitrate to nitrite, which generates a pink azo product fromGriess reagents (sulfanilamide and N-(1-naphthyl)ethylenediamine;Granger et al. 1996; Giovannoni et al. 1997). Samples were handledfollowing the manufacturer’s instructions, except that hemolymph wasdeproteinized with 52 mM zinc sulfate (ZnSO4 ? 7H2O), according toMoshage and Jansen (1998). The azo product absorbance was measuredat 550 nm and nitrite content was quantified per milligram of hemo-lymph protein using a range of sodium nitrite standards.
Protein assay. Samples of pooled hemolymph were sonicated andretained for protein estimation using the Bradford assay (Bradford1975), with bovine serum albumin standards (Sigma).
In vivo inhibition of NADPH oxidase and nitric oxide synthase. Inorder to determine the importance, if any, of the generation of O2
2 and?NO by R. prolixus as part of its anti-trypanosome defenses, twosets of experiments were conducted to inhibit the in vivo enzymaticproduction of these free radicals. The first involved the injection ofthe NADPH oxidase inhibitor N-ethylmaleimide (NEM) into R. pro-lixus with and without T. rangeli hemolymph infections, and the secondexperiment used injections of the inducible NOS inhibitor S-methylisothiourea sulfate (SMT).
The NADPH oxidase inhibitor NEM was freshly dissolved in sterileinsect saline to a concentration of 50 mM, and 2 ml of NEM wasinjected into a group of 25 R. prolixus. Two different sets of insectsreceived injections of NEM combined with either T. rangeli H14 orChoachi strain short epimastigotes (5 3 106 ml21), another two setsreceived injections of T. rangeli H14 or Choachi only, and finally agroup of control insects were injected with insect saline only.
Killed trypanosomes were used as a further negative control and todetermine any potential synergistic activity between the trypanosomesand the inhibitor. In replicates of the above treatments, four furthergroups of R. prolixus were injected using UV-killed T. rangeli Choachior H14 strains, with or without NEM. Trypanosomes were killed by
three 30-min exposures to UV light using an UVP CL-1000 ultravioletcrosslinker (GRI Molecular Biology, UK). Insect mortality was as-sessed at 1 and 5 days postinjection.S-Methyl isothiourea sulfate (Alexis Corp. (UK) Ltd., Notting-hamshire, UK), which acts as an L-arginine analogue (Southan et al.1995), was freshly dissolved in sterile insect saline (see above) to aconcentration of 1 mM, and 2 ml SMT was injected into a group of50 R. prolixus, as described above. A second and third set of 50insects received injections of SMT combined with either T. rangeliH14 or Choachi strain short epimastigotes (5 3 106 ml21), a further
47
two sets received injections of T. rangeli H14 or Choachi only, and,finally, a group of 50 control insects were injected with saline only.In replicates of the above treatments, four further groups of R. prolixuswere injected using UV-killed T. rangeli Choachi or H-14 strains, withor without SMT. As before, insect mortality was assessed at 1 and 5days postinjection.
Statistical analyses. All NBT and PpO assay data were expressedas the mean SOD- or PTU-inhibitable OD change, per milligram ofhemolymph protein, and reported as the mean 6 standard error. The
assay and the NADPH oxidase inhibition assay was analyzed by contin-gency tables using Fisher’s exact test (Motulsky 1995). Relationshipsbetween hemocyte and parasite counts, and NBT assay data, and be-tween the initial 24-h kinetics of PpO and O2
2 generation were analyzedwith Pearson’s correlation test (Motulsky 1995). Differences betweentreatments and between sampling times were considered significantwhen P , 0.05.
RESULTS
Superoxide production. Superoxide activity was mea-sured in vivo in the hemolymph from groups of R. prolixusthat had been injected with two different strains of T. rangeli.In addition, the in vitro response to T. rangeli was measuredin hemolymph from uninjected insects. In all parasite-inocu-lated insects, in vivo O2
2 generation rose after injection,peaked at 24 h, and then fell away to levels comparablewith buffer-injected controls by 72 h (Fig. 1).
Strain H14 short form epimastigote injections resulted inO2
2 generation which was significantly greater at 12 and 24h than that of the buffer-injected control (Figs. 1a and 1e).In addition, O2
2 generated by H14 short form-injected insectswas significantly higher at 24 h than that by insects inocu-lated with the other forms or strains of T. rangeli studied.
H14 strain long form epimastigote injections resulted inO2
2 levels which were significantly greater at 12, 24, and48 h than those of the buffer-injected controls (Figs. 1b and1e), and although significantly lower than H14 short formresponses at 24 h, they were significantly higher at 48 hthan those of any of the other parasite treatments.
Choachi short form injections provoked O22 production
which was significantly greater at 24 h than that in thebuffer-injected controls (Figs. 1c and 1e), though signifi-cantly lower than the response to H14 short forms. Theresponses to Choachi short forms were also significantlyhigher than those to Choachi long forms at 24 and 48 h(Figs. 1c and 1d).
Choachi long form inoculations resulted in the lowest
FIG. 1. Superoxide generation and total hemocyte and parasitecounts in Rhodnius prolixus at various sample times following inocula-tion with Trypanosoma rangeli: (a) H14 strain short epimastigotes; (b)H14 strain long epimastigotes; (c) Choachi strain short epimastigotes;(d) Choachi strain long epimastigotes; or (e) buffer (controls). Superox-ide was measured as the mean SOD-inhibitable NBT reduction/mg ofhemolymph protein 6SE. Cell counts are expressed per milliliter ofhemolymph (log scale); NI, in vitro response of naıve insects. n 5 10.
WHITTEN ET AL.
O22 production responses of any parasite treatment studied,
but the levels were still significantly greater than those ofthe buffer-injected insects at 12 h and the in vitro responsesof naıve insects at 12 and 24 h (Fig. 1d).
There were no significant differences between the O22
responses in buffer-injected control insects at any time (Fig.1e), nor were there any differences between the in vitroresponses of hemolymph from uninjected insects (“NI”) toeither form of T. rangeli strain H14 (Figs. 1a and 1b). How-ever, there was a small but significant rise in in vitro O2
2
generation in response to short and long forms of the Choachistrain, in comparison with the response to buffer alone(P , 0.05; compare Figs. 1c–1e). We also found that theaddition of T. rangeli parasites in vitro to hemolymph derivedfrom parasite-inoculated insects caused no significant en-hancement of O2
2 production (data not shown).Hemocyte and parasite populations. No significant
changes in parasite counts were found in insects injectedwith T. rangeli strain H14 at any time (Figs. 1a and 1b),but in Choachi-inoculated insects, parasite counts increasedsignificantly (Figs. 1c and 1d). Total parasite counts in-creased markedly in Choachi-inoculated insects, particularlyafter 24 h, when O2
2 production began to decline. At thispoint, long form epimastigotes began to increase rapidly andbecame the predominant developmental form by 72 h (Figs.2c and 2d). The number of dividing forms also increasedsignificantly (P , 0.01; Figs. 2c and 2d). In contrast, H14strain parasite counts did not significantly increase over time(compare Figs. 1a and 2a, 1b and 2b), and in both H14-injected groups, the parasite counts dipped as O2
2 levelspeaked at 24 h. Also at this time, in H14 short form-injectedinsects, short forms began to decline as long epimastigotesstarted to increase (Fig. 2a). A large percentage of the H14strain parasites observed in hemolymph from both long-and short form-injected insects were attached to hemocytes,possibly undergoing phagocytosis (Figs. 2a and 2b). Thisphenomenon was also observed in Choachi long form-in-jected insects at 72 h (Fig. 2d).
Although in Choachi-injected insects, dividing forms werepresent after the peak in O2
2 activity (Figs. 2c and 2d), they
48
were only observed in H14 short form-injected insects withinthe first 12 h (Fig. 2a). Despite minor fluctuations, no signifi-cant changes were observed in the total hemocyte counts ofany of the injected insects (Figs. 1a–1e). Figure 3 illustratesformazan deposition within R. prolixus hemocytes, stimu-lated by E. coli LPS. The formazan is clearly visible underbright-field optics.
Prophenoloxidase activation. Significantly elevated
FREE R CTIONS
FIG. 3. (a) Intracellular formazan deposition, indicative of super-oxide production, in R. prolixus hemocytes, in which cells were incu-bated with Escherichia coli LPS and NBT. (b) Unstimulated controlhemocytes, for comparison, in which cells were incubated with NBTalone. Bright field. Bar, 20 mm.
ADICALS IN TRIATOMINE—TRYPANOSOME INTERA
PpO activity was detected in hemolymph from insects in-jected with short form T. rangeli H14, at 24 and 48 h postin-jection (P , 0.01), and with short form T. rangeli Choachiat 24 h (P , 0.05), compared with buffer-injected controls(Fig. 4). This parasite-induced activity was also significantly
49
FIG. 2. Differential parasite counts in R. prolixus at various sampletimes following inoculation with T. rangeli: (a) H14 strain short epimas-tigotes; (b) H14 strain long epimastigotes; (c) Choachi strain shortepimastigotes; (d) Choachi strain long epimastigotes. n 5 10.
ateh fme
FIG. 4. Prophenoloxidase activity in R. prolixus hemolymph inoculwith controls (buffer-injected) and the in vitro response of hemolympChoachi short forms). Prophenoloxidase activity was measured as the6SE. n 5 at least three.
greater than that in uninjected insects. Long form epimasti-gotes of either strain failed to elicit significantly enhancedPpO activity in infected insects, compared with the buffer-injected controls. With the exception of H14 strain longforms, the kinetics of hemolymph PpO activity in responseto all the parasites tested resembled those of O2
2 production,that is, rising to a peak at 24 h and then declining. This wasespecially noticeable in the case of PpO activity in responseto short form H14, which correlated closely with O2
2 genera-tion under the same conditions (R2 5 0.89, P , 0.01; com-pare Figs. 4 and 1a).
The highly variable results for the H14 short form treat-ment at 72 h were attributed to dopachrome/melanin surfacefilms, which interfered with the absorbance readings.
Reactivity of NADPH oxidase p67phox antibodies with R.prolixus hemocytes and plasma. Polyclonal antiserumraised against a sequence of the human NADPH oxidasep67phox protein recognized an immunoreactive band in crudelysate preparations of R. prolixus hemocytes, with an approx-imate molecular mass of 67 kDa. A very faint band of similar
50 WHITTEN ET AL.
mass was observed in R. prolixus plasma (Fig. 5). A strong67-kDa band was also recognized in the human SKN celllysate, used as a positive control, but not in the negativecontrol, in which primary antibody was replaced with rabbitimmune serum.
Nitrite and nitrate production. Combined nitrate andnitrite levels, which represent metabolic products of nitricoxide reactions, were greatly enhanced at 24 and 48 h after
d with different forms and strains of T. rangeli epimastigotes, comparedrom uninjected insects (one representative bar shown, the response toan PTU-inhibitable OD change per milligram of hemolymph protein,
FIG. 5. Western blot for the detection of the NADPH oxidaseprotein p67phox, using anti-peptide anti-p67phox polyclonal antiserum;s, human SKN cell lysate; h, R. prolixus hemocyte lysate; p, R. prolixusplasma. Molecular mass markers (kDa) are indicated on the left.
S
FREE RADICALS IN TRIATOMINE—TRYPANOSOME INTERACTIONinjection in pooled hemolymph samples from T. rangeliH14-injected insects, whereas those in Choachi- and buffer-injected insects remained low (Fig. 6). H14 strain long formepimastigotes stimulated higher levels of hemolymph ni-trates and nitrites than short forms.
In vivo NADPH oxidase inhibition. In order to assessits impact on insect survival, the NADPH oxidase inhibitorNEM was injected into R. prolixus with and without T.rangeli hemolymph infections (Fig. 7). After 1 day, none ofthe injected treatments led to significantly more insectdeaths, compared with any of the other treatments. However,by day 5, insects inoculated with either the H14 or Choachistrain of live T. rangeli (short form epimastigotes) hadsignificantly higher death rates when the hemolymphinfection was combined with NEM injections (72.4%, P ,0.05, and 75.0%, P , 0.01, respectively). The injection ofNEM on its own, or either strain of live parasite on its own,did not cause a significant increase in mortality comparedwith the saline-injected insects, at either 1 or 5 days, nordid the injections involving either strain of the UV-killedtrypanosomes.
In vivo inhibition of nitric oxide synthase. We investi-gated the in vivo effect of the iNOS inhibitor SMT on R.prolixus inoculated with T. rangeli H14 or Choachi (shortforms). Figure 8 shows that after 1 day, insects inoculatedwith either the H14 or Choachi strain of live T. rangelihad significantly higher death rates when the hemolymphinfection was combined with SMT injections (54.7%,P , 0.01, and 40.5%, P , 0.01, respectively). The samephenomenon was observed after 5 days; mortality increasingsignificantly from 45.1 to 85.7% when H14 injections werecombined with SMT treatment (P , 0.0001) and from 43.5to 86.5% when Choachi strain injections were combinedwith SMT (P , 0.001). These mortality rates were alsosignificantly higher than those observed in insects injectedsolely with SMT (both P , 0.01). Compared with saline-injected controls, R. prolixus injected with SMT only had ahigher death rate after 5 days (P , 0.05), but the injectionof either strain of live parasite on its own did not cause asignificant increase in mortality compared with the saline-injected control insects, at either 1 or 5 days, nor did UV-killed trypanosomes on their own. However, R. prolixus
injected with SMT only were less likely to survive after 5days (P , 0.05), and also after 5 days, SMT caused increasesin the mortality of insects co-injected with UV-killed para-sites, which was statistically significant in the case of theChoachi strain of T. rangeli (P , 0.05).The mortalities resulting from combined injections of liveT. rangeli of either strain, plus SMT, were significantlygreater than those among the corresponding groups of insects
51
injected with a combination of UV-killed trypanosomes plusSMT (P , 0.05 for both strains).
DISCUSSION
This study compares superoxide production and its kinet-ics in the hemolymph of R. prolixus inoculated with twodifferent strains and developmental forms of T. rangeli andlinks this activity to other immune mechanisms involvingfree radicals, principally, the nitric oxide and prophenoloxi-dase pathways.
Although several authors have identified O22 or reactive
oxygen species (ROS) production in the hemolymph or he-mocytes of insects, as part of the immune defense (Sun andFaye 1995; Cox-Foster et al. 1998; Fenimore and Cox-Foster1998; Kobayashi 1998; Whitten and Ratcliffe 1999), onlythe work of Nappi et al. (1995) and Nappi and Vass (1998b)have shown ROS production in parasitized insects. The pres-ent study is the first to investigate the immunological roleof hemolymph ROS, RNIs, and PpO in parallel during proto-zoan infections in an insect vector of disease. Experimentsdescribed here show that inoculation of R. prolixus with T.rangeli caused increases in hemolymph O2
2 production. Irre-spective of the developmental form, both the Choachi andH14 strains stimulated in vivo O2
2 production, significantlymore than the controls (buffer-injected and the in vitro re-sponse of naıve insect hemolymph) at 12 h and with all,except Choachi long epimastigotes, at 24 h too.
The injection of H14 short epimastigotes provoked thehighest levels of O2
2 activity in R. prolixus in this study.This experimental group also had the lowest parasitemia,and the most rapid loss of short forms from the hemolymphcoincided with peak O2
2 generation (24 h), while long formspersisted in the hemolymph. A large percentage of the H14parasites were observed attached to hemocytes soon afterinjection, possibly triggering phagocytosis and a pathwayof signaling events that culminated in activation of the respi-ratory burst. In insects inoculated with H14 long forms,
O22 production was lower than that stimulated by H14 shortforms, but significantly higher than that of the controls (at12 and 24 h). At peak O2
2 generation, short forms disappearedfrom the hemolymph, but long forms did not. In both cases,however, the H14 strain parasite numbers did not increasesignificantly over the study time, and invasion of the salivaryglands was not observed, even weeks after inoculation.
Although lower than the response to H14 short forms,
wingrepr
FIG. 6. Total nitrite content of R. prolixus whole hemolymph folloNitrite was determined in pooled samples using Griess reagent andoxide metabolism.
Choachi strain inoculations also provoked significantly en-hanced O2
2 production, but the parasites thrived in the hemo-lymph. As the O2
2 levels began to drop (after 24 h), a largenumber of dividing parasites were observed. There werefewer parasites attached to hemocytes than with the H14strain (only a sharp increase after 48 h in long form infec-tions), indicating perhaps less phagocytosis. As with theH14 strain, Choachi short forms elicited a higher O2
2 re-sponse than the long forms, and by the end of the experiment,short forms began to decline, while the long forms increasedin number (Fig. 2). These results suggest that in both strains,long form epimastigotes are better able to evade or inhibitthe O2
2 response. This finding is in agreement with infectivitystudies conducted in several trypanosome–triatomine sys-tems (e.g., Mello et al. 1995; Feder et al. 1999). The differ-ences in the immune responses to these two strains, andeven between the developmental forms, highlights the im-portance of these factors in studying determinants of infec-tivity and how difficult it is to generalize about the natureof the response. Such variability has been encountered byother workers assessing other aspects of triatomine–trypanosome interactions (e.g., PpO studies in R. prolixusby Gomes et al. 1999).
Preliminary experiments have also shown elevated O22
generation in whole hemolymph from R. prolixus inoculated
with four different strains of T. cruzi, the etiologic agent ofChagas’ disease (Ratcliffe et al., unpublished). As with T.rangeli, peak O22 generation occurred 24 h after inoculation,returning to levels comparable to those of the controls within72 h. All the tested T. cruzi strains failed to divide in vivoand were rapidly cleared from the hemolymph, indicatingthat O2
2 or another ROS may play an important role in thisanti-trypanosome response.
T. rangeli inoculation, compared with that of buffer-injected controls.esents reduced nitrates and nitrites as products of hemolymph nitric
Superoxide can be generated by a host of biologicalsources in addition to the respiratory burst (Davies, 1995).It was thus important to demonstrate putative hemocyticNADPH oxidase, the respiratory burst enzyme, as a sourceof O2
2 in R. prolixus. Some studies on noninsect invertebrateshave demonstrated the presence of NADPH oxidase (Pipe1992; Adema et al. 1993; Smith and Hutton 1995), and ourprevious studies (Whitten and Ratcliffe 1999) showed thatNADPH oxidase inhibitors decreased hemolymph O2
2 pro-duction and bacteriostatic activity in the cockroach Blaberusdiscoidalis. Here, we report that injection of the NADPHoxidase inhibitor N-ethylmaleimide significantly increasedthe mortality rates of T. rangeli-infected insects (bothstrains), strongly suggesting that the respiratory burst has animportant role in anti-trypanosome immunity. Using Westernblotting, we detected a 67-kDa protein from crude R. prolixushemocyte lysates, which was recognized by polyclonal anti-serum raised against a sequence of the human NADPH oxi-dase p67phox protein. The faint band observed in the plasmawas probably a result of slight contamination with debrisfrom hemocytes that broke down during sample preparation.It is not surprising to find an insect protein immunologicallyrelated to human NADPH oxidase, since similar proteinshave been identified in a wide range of organisms, fromplants (Dwyer et al. 1996) to numerous mammals (Hitt and
52 WHITTEN ET AL.
Kleinberg 1996). Combined with the O22 assay data, the
present results provide compelling evidence for the existenceof a respiratory burst NADPH oxidase system in R. pro-lixus hemolymph.
The present study has also shown significantly increasedPpO activity associated with T. rangeli short form infectionsusing either the Choachi or H14 strains. However, longepimastigotes from either strain failed to elicit a significantly
lled(N-e
FIG. 7. Mortality rates in R. prolixus injected with live or UV-kieffect of the concurrent injection of the NADPH oxidase inhibitor NEM
enhanced response. The fact that in H14 short epimastigoteinfections, PpO activity correlated with the kinetics of O2
2
formation, may indicate that products of these pathwaysinteract. For example, O2
2 can produce highly toxic ROS byreacting with dihydroquinones (Nappi and Vass 1993), butequally, melanin could serve to protect the host from thewidespread dissemination of free radicals, by scavengingO2
2 (Korytowski et al. 1986; Professor A. J. Nappi, per-sonal communication).
It is unlikely that O22 alone would be responsible for any
anti-parasitic effects in the systems studied here, as this freeradical is only mildly oxidizing compared with some of itsROS products. Often, toxicity associated with O2
2 is ascribedto hydroxyl radicals, which may, for example, be generatedby the interaction of O2
2 with hydrogen peroxide, via transi-tion metal-catalyzed ion transfer (Haber-Weiss reaction;
Cheeseman and Slater 1993) or by the combination of ?NOwith hydrogen peroxide (Nappi and Vass 1998a).To assess whether ?NO generative activities could operateduring T. rangeli infection in R. prolixus, we measured RNIbreakdown products, nitrate and nitrite, in inoculated hemo-lymph. These products rose in H14 strain-injected insects,in response to both long and short epimastigotes at 24 and 48h, but Choachi strain-injected hemolymph failed to produce
T. rangeli strains H14 or Choachi (short form epimastigotes) and thethylmaleimide). The minimum number of insects per treatment was 20.
levels greater than the buffer-injected controls. Additionally,preliminary experiments have shown that the hemolymphtotal nitrite content of T. cruzi-inoculated R. prolixus wasenhanced at 24 and 48 h after injection, and T. cruzi did notsurvive in the hemolymph (Whitten et al., unpublished).These elevated hemolymph nitrite levels, suggestive of in-creased ?NO production, were at least equal to those mea-sured in insects injected with T. rangeli H14 strain. Althoughin this paper, we did not attempt to distinguish betweenintracellular and extracellular production of ?NO and RNIs,it is known that in mammals, both these mechanisms areutilized against trypanosomes (e.g., Gobert et al. 1998;Rottenberg et al. 1996). In the present study, the H14 strainof T. rangeli was more often observed attached to hemo-cytes than the Choachi strain, and it may be that contactwith hemocytes and intracellular ROS/RNI production is
FREE RADICALS IN TRIATOMINE—TRYPANOSOME INTERACTIONS 53
more important in terms of anti-trypanosome defenses inR. prolixus.
We next determined the impact of a NOS inhibitor in vivoon T. rangeli H14 hemolymph infections using S-methylisothiourea sulfate, which selectively inhibits the inducibleNOS isoform in mammals. Our initial intention was to com-pare hemolymph parasite counts; however, it was soon evi-dent that many of the infected insects treated with SMT had
d T.m n
development. Prophenoloxidase activity was also up-regu-
FIG. 8. Mortality rates in R. prolixus injected with live or UV-killeof the iNOS inhibitor SMT (S-methyl isothiourea sulfate). The minimu
died, and the experiment was adapted to quantify insectmortality. When combined with live T. rangeli H14 andChoachi strain parasites, this inhibitor caused a significantincrease in mortality among R. prolixus at 1 and 5 days afterinoculation, suggesting that ?NO generation may play animportant role in the host’s immune response. The mortalitiesresulting from combined injections of SMT with viable T.rangeli were significantly higher than those of the corres-ponding insect groups receiving SMT plus UV-killed trypa-nosomes. The effect with the Choachi strain is surprising asit contradicts the low levels of ?NO products measured withthe Griess assay. Perhaps only a very low production of ?NOis sufficient to exert anti-parasitic effects, for example, by
?NO acting as a messenger molecule (Stamler 1994). Inter-estingly, SMT alone significantly increased insect mortalityat day 5, possibly as a result of toxicity or possibly becausenitric oxide inhibition also impacts other important physio-logical functions. SMT was chosen in preference to otheravailable NOS inhibitors primarily because of its high selec-tivity toward the inducible isoform of the enzyme. However,it would be interesting to compare the above results withrangeli H14 or Choachi strain short form epimastigotes and the effectumber of insects per treatment was 20.
those of other NOS inhibitors, such as L-NAME (Luckhartet al. 1998), which may have lower toxicity in R. prolixus.
In summary, we have used a combination of biochemicaland immunological studies to provide evidence for O2
2 gener-ation and the respiratory burst enzyme NADPH oxidase inR. prolixus hemolymph and hemocytes. This O2
2 responsewas inducible by T. rangeli hemolymph infection, thoughthe magnitude varied with the parasite strain and stage of
54 WHITTEN ET AL.
lated in response to T. rangeli infection and its kineticscorrelated with O2
2 generation. In addition, products of RNIswere elevated in response to one strain of T. rangeli, andinhibitor studies suggested a role for NOS activity as a
determinant of T. rangeli infectivity. This NOS activity, andits possible interaction with the respiratory burst system,clearly requires further detailed investigation.ACKNOWLEDGMENTS
We are grateful to the Royal Society for an equipment grant and tothe Wellcome Trust for a travel grant, awarded to N.A.R., and to Mrs.
FREE RADICALS IN TRIATOMINE—TRYPANOSOME INTERACTIONS
Christine Day and Mr. Ian Tew for assistance in rearing R. prolixus.C.B.M., P.A., and E.S.G. are CNPq research fellows. This work wassupported by grants from CNPq, FIOCRUZ, FAPERJ, FINEP, andPADC. M.M.A.W. was funded by a postgraduate studentship from theUniversity of Wales. The authors also thank Professor A. J. Nappi forhis helpful comments.
REFERENCES
Adema, C. M., Van Deutekom-Mulder, E. C., Van Der Knaap,W. P. W., and Sminia, T. 1993. NADPH-oxidase activity: The proba-ble source of reactive oxygen intermediate generation in hemocytesof the gastropod Lymnaea stagnalis. Journal of Leukocyte Biology54, 379–383.
Azambuja, P., and Garcia, E. S. 1997. Care and maintenance of triato-mine colonies. In “Molecular Biology of Insect Disease Vectors: AMethods Manual” (J. M. Crampton, C. B. Beard, and C. Louis,Eds.), pp. 56–64. Chapman & Hall.
Babior, B. M., Kipnes, R. S., and Curnutte, R. S. 1973. Biologicaldefense mechanisms. The production by leukocytes of superoxide,a potential bactericidal agent. The Journal of Clinical Investigation52, 741–744.
Bradford, M. M. 1975. A rapid and sensitive method for the quantifica-tion of microgram quantities of protein utilizing the principle ofprotein-dye binding. Analytical Biochemistry 72, 248–254.
Brisby, R. H. 1990. Pulse radioloysis, free radicals and antiparasiticdrugs. Parasitology Today 6(3), 84–88.
Cheeseman, K. H., and Slater, T. F. 1993. An introduction to freeradical biochemistry. British Medical Bulletin 49(3), 481–493.
Conte, A., and Ottaviani, E. 1995. Nitric oxide synthase activity inmolluscan hemocytes. FEBS Letters 365, 120–124.
Cox-Foster, D. L., Fenimore, J. A., and Lazzaro, B. 1998. Biochemistryof oxidative free radical production during innate immunity by insecthemocytes. Free Radical Biology and Medicine 25(Suppl. 1), S44.
Davies, K. J. A. 1995. Oxidative stress: The paradox of aerobic life.Biochemical Society Symposium 61, 1–31.
Dimopoulos, G., Seeley, D., Wolf, A., and Kafatos, F. C. 1998. Malariainfection of the mosquito Anopheles gambiae activates immune-responsive genes during critical transition stages of the parasite lifecycle. The EMBO Journal 17(21), 6115–6123.
Docampo, R., Casellas, A. M., Madeira, E. D., Cardona, R. L., Moreno,S. N. J., and Mason, R. P. 1983. Oxygen-derived radicals fromTrypanosoma cruzi-stimulated human neutrophils. FEBS Letters
155(1), 25–30.Dwyer, S. C., Legendre, L., Low, P. S., and Leto, T. L. 1996. Plantand human neutrophil oxidative burst complexes contain immuno-logically related proteins. Biochimica et Biophysica Acta 1289,231–237.
Fang, F. C. 1997. Mechanisms of nitric oxide-related antimicrobialactivity. Journal of Clinical Investigation 100(12 Suppl.), S43–S50.
Feder, D., Gomes, S. A. O., Garcia, E. S., and Azambuja, P. 1999.
55
Metalloproteases in Trypanosoma rangeli-infected Rhodnius pro-lixus. Memorias do Instituto Oswaldo Cruz 94(6), 771–777.
Fenimore, J. A., and Cox-Foster, D. L. 1998. Oxidative free radicalproduction during encapsulation in Manduca sexta. In “Abstractsfrom the Third International Symposium on Molecular InsectScience,” Snowbird, Utah, p. 33.
Franchini, A., Fontanili, P., and Ottaviani, E. 1995. Invertebrate immu-nocytes: Relationship between phagocytosis and nitric oxide produc-tion. Comparative Biochemistry and Physiology 110B(2), 403–407.
Gabig, T. G., and Babior, B. M. 1981. The killing of pathogens byphagocytes. Annual Review of Medicine 32, 313–326.
Garcia, E. S., and Azambuja, P. 1997. Infection of triatomines withTrypanosoma cruzi. In “Molecular Biology of Insect Disease Vectors:A Methods Manual” (J. M. Crampton, C. B. Beard, and C. Louis,Eds.), pp. 146–155. Chapman & Hall.
Giovannoni, G., Land, J. M., Keir, G., Thompson, E. J., and Heales,S. J. R. 1997. Adaptation of the nitrate reductase and Griess reactionmethods for the measurement of serum nitrate plus nitrite levels.Annals of Clinical Biochemistry 34, 193–198.
Gobert, A. P., Semballa, S., Daulouede, S., Lesthelle, S., Taxile, M.,Veyret, B., and Vincendeau, P. 1998. Murine macrophages use oxy-gen- and nitric oxide-dependent mechanisms to synthesize S-nitroso-albumin and to kill extracellular trypanosomes. Infection and Immu-nity 66(9), 4068–4072.
Gomes, S. A. O., Feder, D., Thomas, N. E. S., Garcia, E. S., andAzambuja, P. 1999. Rhodnius prolixus infected with Trypanosomarangeli: In vivo and in vitro experiments. Journal of InvertebratePathology 73, 289–293.
Granger, D. L., Taintor, R. R., Boockvar, K. S., and Hibbs, J. B. Jr.1996. Measurement of nitrate and nitrite in biological samples usingnitrate reductase and Griess reaction. Methods in Enzymology268, 142–151.
Gregorio, E. A., and Ratcliffe, N. A. 1991. The prophenoloxidasesystem and in vitro interaction of Trypanosoma rangeli with Rhod-nius prolixus and Triatoma infestans haemolymph. Parasite Immu-nology 13, 551–564.
Hecker, H., Schwarzenbach, M. and Rudin, W. 1990. Developmentand interactions of Trypanosoma rangeli in and with the reduviidbug Rhodnius prolixus. Parasitology Research 76, 311–318.
Hitt, N. D., and Kleinberg, M. E. 1996. Identification of neutrophilNADPH oxidase proteins gp91-phox, p22-phox, p67-phox, and p47-phox in mammalian species. American Journal of VeterinaryResearch 57(5), 672–676.
Hughes, H. P. A. 1988. Oxidative killing of intracellular parasitesmediated by macrophages. Parasitology Today 4(12), 340–347.
Kobayashi, A. 1998. Insect defence molecules and reactive oxygen.Developmental and Comparative Immunology 22(1), 129.
Korytowski, W., Kalanaraman, B., Menon, I. A., Sarna, T., and Sealy,R. C. 1986. Reaction of superoxide anions with melanins: Electronspin resonance and spin trapping studies. Biochimica et BiophysicaActa 882, 145–153.
Laemmli, U. K. 1970. Cleavage of structural proteins during the assem-bly of the head of bacteriophage T4. Nature 227, 680–685.
56
Leonard, C., Soderhall, K., and Ratcliffe, N. A. 1985. Studies onprophenoloxidase and protease activity of Blaberus craniifer haemo-cytes. Insect Biochemistry 15(6), 803–810.
Luckhart, S., Vodovotz, Y., Cui, L., and Rosenberg, R. 1998. Themosquito Anopheles stephensi limits malaria parasite developmentwith inducible synthesis of nitric oxide. Proceedings of the NationalAcademy of Sciences USA 95, 5700–5705.
Luckhart, S., and Rosenberg, R. 1999. Gene structure and polymor-phism of an invertebrate nitric oxide synthase gene. Gene 232, 25–34.
Martınez, A. 1995. Nitric oxide synthase in invertebrates. Histochemi-cal Journal 27, 770–776.
Mello, C. B., Garcia, E. S., Ratcliffe, N. A., and Azambuja, P. 1995.Trypanosoma cruzi and Trypanosoma rangeli: Interplay with hemo-lymph components of Rhodnius prolixus. Journal of InvertebratePathology 65, 261–268.
Moncada, S., Palmer, R. M. J., and Higgs, E. A. 1991. Nitric oxide:Physiology, pathophysiology and pharmacology. Pharmacology Re-views 43(2), 109–142.
Moshage, H., and Jansen, P. L. M. 1998. Adaption of the nitrate andGriess reaction methods for the measurement of serum nitrate plusnitrate levels. Annals of Clinical Biochemistry 35, 154–155.
Motulsky, H. 1995. In “Intuititve Biostatistics,” various chapters.Oxford Univ. Press, New York.
Nappi, A. J., and Vass, E. 1993. Melanogenesis and the generation ofcytotoxic molecules during insect cellular immune reactions. Pig-ment and Cell Research 6, 117–126.
Nappi, A. J., Vass, E., Frey, F., and Carton, Y. 1995. Superoxideanion generation in Drosophila during the melanotic encapsulationof parasites. European Journal of Cell Biology 68, 450–456.
Nappi, A. J., and Vass, E. 1998a. Hydroxyl radical formation resultingfrom the interaction of nitric oxide and hydrogen peroxide. Biochim-ica et Biophysica Acta 1380, 56–63.
Nappi, A. J., and Vass, E. 1998b. Hydrogen peroxide production inimmune-reactive Drosophila melanogaster. Journal of Parasitology84(6), 1150–1157.
Nathan, C., Nogueira, N., Juangbhanich, C., Ellis, J., and Cohn, Z.1979. Activation of macrophages in vivo and in vitro. Correlationbetween hydrogen peroxide release and killing of Trypanosomacruzi. Journal of Experimental Medicine 149, 1056–1068.
Pereira, M. E. A., Andrade, A. F. B., and Ribeiro, J. M. C. 1981. Lectinsof distinct specificity in Rhodnius prolixus interact selectively withTrypanosoma cruzi. Science 211, 597–600.
Pick, E., and Mizel, D. 1981. Rapid microassays for the measurementof superoxide and hydrogen peroxide production by macrophagesin culture using an automatic enzyme immunoassay reader. Journal
of Immunological Methods 16, 111–122.Pipe, R. K. 1992. Generation of reactive oxygen metabolites by thehaemocytes of the mussel Mytilus edulis. Developmental and Com-parative Immunology 16, 111–122.
Ratcliffe, N. A., Nigam, Y., Mello, C. B., Garcia, E. S., and Azambuja, P.1996. Trypanosoma cruzi and erythrocyte agglutinins: A comparativestudy of occurrence and properties in the gut and hemolymph ofRhodnius prolixus. Experimental Parasitology 83, 83–93.
WHITTEN ET AL.
Rottenberg, M. E., Castanos-Velez, E., De Mesquita, R., Laguardia,O. G., Biberfeld, P., and Orn, A. 1996. Intracellular colocalizationof Trypanosoma cruzi and inducible nitric oxide synthase (iNOS):Evidence for dual pathway of iNOS induction. European Journalof Immunology 26, 3203–3213.
Segura-Aguilar, J. 1993. A new direct method for determining superox-ide dismutase activity by measuring hydrogen peroxide formation.Chemico-Biological Interactions 86, 69–78.
Slepneva, I. A., Glupov, V. V., Sergeeva, S. V., and Khramtsov, V. V.1999. EPR detection of reactive oxygen species in hemolymph ofGalleria mellonella and Dendrolimus superans sibiricus (Lepidop-tera) larvae. Biochemical and Biophysical Research Communications264, 212–215.
Smith, V. J., and Hutton, D. M. C. 1995. Measurement of extracellularreactive oxygen species in anthozoans and other marine invertebratesby cytochrome c reduction. In “Techniques in Fish Immunology.Immunology and Pathology of Aquatic Invertebrates” (J. S. Stolen,T. C. Fletcher, S. A. Smith, J. T. Zelikoff, S. L. Kaattari, R. S.Anderson, K. Soderhall, and B. A. Weeks-Perkins, Eds.), pp. 27–35.SOS, Fair Haven, U.S.A.
Southan, G. J., Szabo, C., and Thiemermann, C. 1995. Isothioureas:Potent inhibitors of nitric oxide synthases with variable isoformselectivity. British Journal of Pharmacology 114, 510–516.
Stamler, J. S. 1994. Redox signalling: Nitrosylation and related targetinteractions of nitric oxide. Cell 78, 931–936.
Sun, S. C., and Faye, I. 1995. Transcription of immune genes in thegiant silkmoth, Hyalophora cecropia, is augmented by H2O2 anddiminished by thiol reagents. European Journal of Biochemistry231, 93–98.
Takle, G. B. 1988. Studies on the cellular immune response of insectstoward the insect pathogen Trypanosoma rangeli. Journal of Inverte-brate Pathology 51, 64–72.
Vespa, G. N. R., Cunha, F. Q., and Silva, J. S. 1994. Nitric oxide isinvolved in control of Trypanosoma cruzi-induced parasitemia anddirectly kills the parasite in vitro. Infection and Immunity 62(11),5177–5182.
Watkins, R. 1971. Histology of Rhodnius prolixus infected with Trypa-nosoma rangeli. Journal of Invertebrate Pathology 17, 59–66.
Weiske, J., and Wiesner, A. 1999. Stimulation of NO synthase activityin the immune-competent Lepidopteran Estigmene acraea hemocyteline. Nitric Oxide: Biology and Chemistry 3(2), 123–131.
Whitten, M. M. A., and Ratcliffe, N. A. 1999. In vitro superoxideactivity in the haemolymph of the West Indian leaf cockroach, Blab-erus discoidalis. Journal of Insect Physiology 45, 667–675.
Wientjes, F. B., and Segal, A. W. 1995. NADPH oxidase and therespiratory burst. Seminars in Cell Biology 6, 357–365.
Wiesner, A. 1991. Induction of immunity by latex beads and by hemo-lymph transfer in Galleria mellonella. Developmental and Compara-tive Immunology 15, 241–250.
Yuda, M., Hirai, M., Miura, K., Matsumura, H., Ando, K., and Chinzei,Y. 1996. cDNA cloning, expression and characterization of nitric-oxide synthase from the salivary glands of the blood-sucking insectsRhodnius prolixus. European Journal of Biochemistry 242, 807–812.
Zeledon, R. 1987. Life cycle of Trypanosoma cruzi in the insect vector.
FREE RADICALS IN TRIATOMINE—TRYPANOSOME INTERACTIONS 57
In “Chagas Disease Vectors (Anatomical and Physiological As- Nitric Oxide, Vol. 2, Enzymology, Biochemistry and Immunology”
pects)” (R. R. Brenner and A. M. Stoka, Eds.), pp. 59–75. CRCPress, Boca Raton, Florida.Zhu, L., Gunn, C., and Beckman, J. S. 1992. Peroxynitrite, the productof nitric oxide and superoxide, is bactericidal. In “The Biology of
(S. Moncada, M. A. Marletta, J. B. Hibbs Jr., and E. A. Higgs, Eds.),pp. 245–247. Portland Press, London.
Received 10 October 2000; accepted with revision 1 March 2001