InternationalJounwlfor f’urasitolo~ hinted in Great Brifoin.
Vol. 16. No. 6, pp. 6
0020-7519/86 $3.00+ 0.00 t’qamon
1586 Ausrralian Society for l’arasitology.
THE ROLE OF HOST GENERATED FREE RADICALS IN HELMINTH INFECTIONS: NIPPOSTRONGYLUS BRASILIENSIS AND NEMATOSPIROIDES DUBIUS COMPARED N. C. SMITHand C. BRYANT Department of Zoology, Australian National University, GPO Box 4, Canberra 2601, Australia (Received 16 September 1985) Abstract-hmi N. C. and BRYANT C. 1986. The role of host generated free radicals in helminth infections: Nippostrongylus brasiliensis and Nematospiroides dubius compared. International Journalfor Parasitology 16: 6 17-622. The possibility that free radicals are involved in the expulsion of Nippostrongyhs brasiliensis from the small intestine of its mice host was explored by comparing the susceptibilities to free radicals, and levels of protective enzymes, of adult N. brasiliensis and Nematospiroides dubius, a closely related intestinal parasite of mice. Nippostrongyhs brasiliensis was markedly more susceptible to in vitro free radical damage than N. dub&. The difference in susceptibility is probably related to differences in enzymatic protection against free oxygen radicals as N. dubius had roughly twice as much superoxide dismutase, about 3 times as much cat&se and about 4 times as muchglutathione reductase as N. brasiliensis. This result may indicate that N. dubius persists in the rodent small intestine, whilst N. brasiliensis is spontaneously expelled, because of a more efficient enzymatic defence system against host-generated free
oxygen radicals. INDEX KEY WORDS: Parasitic helminths; Nippostrongylus brasiliensis; Nematospiroides dubius; free radicals; immune response; superoxide dismutase; catalase; glutathione reductase; self-cure phenomenon.
monocytes, macrophages, eosinophils, neutrophils and basophils have all been implicated in the expulsion of NippostrongyIus brasiliensis from the small intestine of its rodent host (Miller, 1971; Mackenzie, Jungery, Taylor & Ogilvie, 198 1; Ogilvie, Hesketh & Rose, 1978). These leukocytes appear to mediate the characteristic inflammatory response of the host’s small intestine to N. brusiliensis, thus raising the possibility that free oxygen radicals are involved in the rejection mechanism. Though the effects of free radicals on parasitic helminths are virtually unexplored, it has been demonstrated (Bass & Szejda, 1979) that leukocyte-generated free oxygen radicals are involved in the killing of larval stages of Trichinella spiralis. Kazura & Meshnick (1984) showed that the migratory larvae of T. spiralis were more susceptible to free radical damage than adult or muscle stage larvae because of deficiencies in levels of protective enzymes such as superoxide dismutase and glutathione peroxidase. Together with glutathione reductase and catalase, these enzymes form a primary biological defence against free oxygen radicals. The hypothesis that free oxygen radicals are involved in the expulsion of N. brasiliensis was tested by comparing the susceptibilities of N. brasiliensis and Nematospiroides dubius I==Heligmosomoides polygyrus] to free radicals in vitro. Nippostrongvlus MAST
and Nematospiroides are closely related genera, making up two of the three genera of the superfamily Heligmosomatoidea (Cheng, 1973). Both are intestinal parasites of rodents and are presumably initially exposed to similar non-specific immune responses. However, unlike N. brusiliensis, which is expelled from the rat in lo-12 days, N. dubius persists in the small intestine of the mouse for some months. The possibility that this difference in survival time is related to levels of enzymatic protection against free oxygen radicals was investigated.
MATERIALS AND METHODS Parasites. Female Wistar rats were infected with about 4000 third stage N. brasiliensis larvae by subcutaneous injection. Rats were killed 7 days after infection, the small intestines removed and split open and adult worms collected with a Baermann apparatus. Female Quackenbush mice were infected orally with about 200 third stage N. dubius larvae. Mice were killed 3 weeks after infection and the adult worms collected in the same manner as N. brasiliensis. Rats and mice were killed with an overdose of anaesthetic ether. Evaluation of effects of oxidants/antioxidants. Incubations of adult worms were carried out in 1.0 ml of Earle’s Balanced Salts Solution, comprising NaCl (120 mM), KC1 (5 mu), MgSO, * 7Hz0 (0.4 mM), NaHCO, (11 mM), CaClz * 2Hz0 (1.8 mu) and MOPS0 (S-[N-morpholinol-2hydroxypropane sulfonic acid, 20 mM) plus 50 mM glucose with or without various oxidants and antioxidants at 37°C.
About 25 worms were used for each incubation. All incubations were done in triplicate at least. At the end of the incubations viability was determined by motility. The combined effects of hydrogen peroxide (HZOz) and superoxide radical (02J and possibly hydroxyl radical (* OH) were tested by inclusion of 0.2 mM acetaldehyde and 10 milliunits ml-’ xanthine oxidase in the incubation mixture (Kazura & Meshnick, 1984). Incubations with acetaldehyde-xanthine oxidase were carried out for 12 h. Various antioxidants were included in the acetaldehydexanthine oxidase system to confirm the role of free radicals in mediating damage to N. brasiliensis. Catalase (5000 units ml-‘) was added to eliminate any H,OZ generated. Superoxide dismutase (10 ,ug ml-‘) was included to eliminate O,-. Non-specific antioxidants were also incorporated into the oxidant generating system. These were ihiourea (40 mM), mannitol(10 mM)-& 2,5-dimethylfuran (50 mM). The iron-chelator 2,3-dihvroxv benzoic acid (IO mM)‘was used to assess the imp&a&e of . OH in causing parasite death. The free radical generator, lbutylhydroperoxide, was included in the incubation mixture of N. brusiliensis at a concentration of 10 mM for 20 h. The protective effects of a&oxidants against f-butylhydroperoxide were also tested. Determination of protective enzyme levels. Worms were homogenized in 0.1% Triton X-100 with a ground glass/ ground glass homogeniser. Crude enzyme extracts were centrifuged briefly at high speed in a microcentrifuge. The resulting supernatant was used for enzyme assays and for protein determinations by the Coomassie Brilliant Blue technique of Bradford (1976). Superoxide dismutase activity was determined by the method of SaIin & McCord (1974), and that of catalase was based on the method of Ganschow & Schimkle (1969). The method of Hopkins & Tudhope (1973) was used to assay glutathione peroxidase activity except that 90 mM HEPES buffer (pH 7.0) and 0.7 mM tbutylhydroperoxide were substituted for phosphate buffer
and H,O, respectively (Kazura & Meshnik, 1984). Glutathione reductase was assayed as described in Bergmeyer (1974). Oxygen uptake. The oxygen uptake of about 25 mg (wet wt) of live adult N. brasiliensis or N. dubius in 2 ml of a 50/ 50 mixture of oxygen electrode medium and Earle’s Balanced Salts Solution was determined in a Clarke-type oxygen electrode linked to a Pharmacia one-channel
recorder. Oxygen electrode medium consisted of sucrose (1 M); KC1 (0.5 M); TRIS (0.15 M); KH?PO, (0.1 M); MgSO, .7H,O (20 mM) and EDTA (10 mM). Oxygen uptake was expressed as nmol O2 min-’ gg’ wet wt. Reagents. All chemical reagents were analytical grade. All enzymes and cofactors were obtained from BoehringerMannheim; HEPES, MOPSO, thiourea and Coomassie Brilliant Blue G-250 from Sigma; TRIS from CalbiochemBehring; and 2,5-dimethylfuran and 2,3_dihydroxybenzoic acid from Aldrich. Statistical analyses. First, variances were analyzed using either Bartlett’s test for the equality of several variances or a simple F-test. Differences in the means were tested by the appropriate Student’s t-test depending on whether or not variances could be assumed to be equal.
Effect of oxidants and antioxidants on adult Nippostrongylus brasiliensis and Nematospiroides dubius in vitro Both N. brasiliensis and N. dubius were killed by the acetaldehyde-xanthine oxidase enzyme system but the susceptibility of N. brasiliensis was significantly higher. Ninety-three per cent of adult N. brasiliensis were killed in 12 h by acetaldehydexanthine oxidase whilst only 28% of adult N. dubius were killed in the same time. Similarly, 43% of adult N. brasiliensis were killed by t-butylhydroperoxide in 20 h whilst only 24% of adult N. dubius died (Table 1). The inclusion of various antioxidants in the mixture containing xanthine oxidase and acetaldehyde generally improved survival of N. brasiliensis (Table 2). Superoxide dismutase ,reduced the death rate from 93 to 7%, but catalase was without effect. The non-specific antioxidants thiourea and 25 dimethylfuran both significantly reduced the lethal effect of the acetaldehyde-xanthine oxidase system (from 93 to 9 and lO%, respectively). Mannitol had a small but significant effect. The iron-chelator 2,3_dihydroxybenzoic acid was also an effective
TABLE ~-EFFECT OF OXIDANTS ON ADULT Nippostrongylus brasiliensis AND Nematospiroides dubius in vitro % Killed (mean f s.D.; n = 5) Oxidant
None (control) t-Butylhydroperoxide Acetaldehyde Xanthine oxidase Acetaldehydexanthine oxidase
10mM 0.2 mM 10 m units ml-’ 0.2 rnM 10 m units ml-’
Of0 43f 15h o+o 1t1
Of0 2457” o+o Ok0
93 + 5”
28 + 8”
Twenty-five parasites were incubated at 37 ‘C with or without the various reagents in 1 ml of Earle’s Balanced Salts Solution for 12 h (with acetaldehyde-xanthine oxidase) or 20 h (with t-butylhydroperoxide and control groups). Viability was determined by motility. r.h Pairs of results significantly different at the levels P< 0.0005 and P< 0.025 respectively.
Free radicals and hel~inth infections TABLE
Nippostrongyhs brasiliensisin vitro % Killed (mean +
93*5 7+6” 93 + 8 9+ 5” 84k3b 10 + 10a 25k11s
None (control) Superoxide dismutase (10 pg rn-‘) Catalase (5000 units ml-‘) Thiourea (40 mM) Mannitol(l0 trm) 2,5-Dimethylfuran (SO mn) 2,3-Dihydroxybenzoic acid (10 mM) 2,3-Dihydroxy~~oic acid (10 mu) preincubated for 24 h with 100 mM Fez+ and 100 mn Fe)+
30 + 20”
Twenty-five parasites were incubated with or without the various antioxidants in 1 ml of Earl’s Balanced Salts Solution containing 0.2 rnn acetaldehyde and 10 m units ml-’ xanthine oxidase. Incubations were for 12 h at 37°C. Viability was determined by motility. ab Si~fi~~tIy different from the control at the levels P < 0.0005 and P < 0.005 respectively.
protectant, reducing the death rate of N. brasiliensis from 93 to 25%. This effect was not abolished by saturation of 2,3-dihydro~be~oic acid with ferrous and ferric ions before in~uba~on. The ~tio~dants by themselves did not affect worm survival. t-Butylhydroperoxide was responsible for the death of 43% of N. brasiliensis within 20 h of in vitro incubation. Superoxide dismutase did not alleviate this effect but catalase, thiourea, mannitol, 2,5dimethyl~r~ and 2,3-dihydro~~nzoic acid all si~i~~~~y reduced it from 43 to 2 1, 11,2 1,16 and 10% respectively (Table 3). The antioxidants by themselves had no effect on worm survival.
(Table 4). Nematospiroides dubius had about twice as much superoxide dismutase activity (31.63 f 8.92 units mg-’ protein, compared with 15.34 k 3.45 units mg-’ protein), 3-4 times more catalase activity (3.78 + 0.17 units mg-’ protein compared with 1.13 f 0.23 unit mg-’ protein) and almost 4 times more glutathione reductase activity (0.166 f 0.05 compared with 0.043 k 0.029units mg-’ protein) than N. brasiliensis. Glutathione peroxidase activities were not si~i~c~dy different (0.029 + 0.003 units mg-’ protein for N. br~iZie~~ and 0.023 F 0.006 for N. dubius).
Protective enzyme levels in Nippostrongylus
brasiliensis and Nematospiroides dubius The specific: activities of protective enzymes in adult N. brasil~~~ and N. dubius were determined
Nippostrongylus brasiliensis had an average oxygen uptake rate of 300 rt 62 nmol 0, min-’ g-’ wet wt (mean f s.D., n = 3) compared with an oxygen
Oxygen uptake in Nippostrongylus
Nippostrongyhs brasiliensisin vitro ~tio~d~t None (control) Superoxide dismutase (10 ,u g ml-‘) Catalase (5000 units ml-l) Thiourea (40 mn) Mannitol(l0 mM> 2,5-D~e~yI~r~ (50 mM) 2,3-Dihydroxy~~oic acid (10 mu)
% Killed (mean z!zs.D.; n
43f15 36j: 11 21rt5b 11rf:Y 21z!~7~ 16*6a lOrt6”
Twenty-five parasites were incubated with or without the various antioxidants in 1 ml of Earle’s Balanced Salts Solution containing 10 mn t-butylhydroperoxide for 20 hat 37’C. Viability was determined by motility. a.bSigniticantly different from the control at the levels Pi 0.0025 and PC 0.01 respectively.
N. C. SMITH and C. RRYAPU TABI 4-seEclFIc ACTlVlTlESOF PROTECTIVE LNZYMESIIU Nippostrongylus brasiliensis ANDNematospiroides dubiu Enzyme (s.a. in units mg-’ protein) mean f S.D. Superoxide
15.34 f 3.45,’ (n = 5) 1.13 f 0.23h (n = 5) 0.029 k 0.003 (n = 5) 0.043 f 0.029h (n = 5)
Nemutospiroides dubius 31.63 (n 3.78 (n 0.023 (n 0.166 (n
f = f = f = * =
8.92 3) 0.17 3) 0.006 3) 0.050 5)
1 unit of superoxide dismutase activity is defined as that amount of enzyme required to inhibit the rate of reduction of cytochrome c by 50%. 1 unit of catalase activity is defined as that amount of enzyme necessary to decompose 50% of 20 rnM HzOz in 1 min. 1 unit of glutathione peroxidase activity is defined as that amount of enzyme necessary to oxidise 1 ,u mol of reduced glutathione in I min.
1 unit of glutathione reductase activity is defined as that amount of enzyme necessary to reduce 1 ,LL mol of oxidised glutathione in 1 min. ;’ Significantly different from superoxide dismutase activity of N. dubius at the level P< 0.005 (Student’s t-test). h Significantly different from catalase or glutathione reductase activity of N. dub&s at the level P< 0.0005 (Student’s t-test).
uptake rate of 180 f 10 nmol0, (n = 3) for N. &Gus.
min-’ gg’ wet wt
DISCUSSION Kazura & Meshnick (1984) found a good correlation between the susceptibility of developmental stages of Trichinella spiralis to the immune system of the host and the levels of enzymes protective against free radicals possessed by those stages. Thus, the most susceptible migratory larvae of T. spiralis are killed by the acetaldehyde-xanthine oxidase enzyme system whilst adults and muscle stage larvae are resistant to oxidant damage. The levels of glutathione peroxidase and superoxide dismutase are considerably lower in the migratory larvae than in other stages of T. spiralis. Kazura & Meshnick (1984) argue that as immunity to the early stage of the parasite is a cell mediated response and that as white cells generate free radicals as part of their offensive armoury, the greater enzyme protection achieved by later stages of T. spiralis contributes significantly to their survival within the host. The results presented here show that Nematospiroides dubius, a persistent intestinal parasite of mice, is more resistant to free radical damage in vitro than Nippostronsylus brasiliensis, which is expelled from the intestine of its rat host lo-12 days after infection. This susceptibility to oxidant damage is probably related to levels of protective enzymes as N. dubius has roughly twice as much superoxide dismutase and 3-4 times more catalase
than N. brasiliensis. Although there is no difference in the activities of glutathione peroxidase in the two genera, this may not matter as catalase has similar protective properties to glutathione peroxidase. Furthermore, N. dubius has considerably higher levels of activity of glutathione reductase, the enzyme responsible for ensuring maintenance of adequate levels of reduced glutathione. This is important on two counts: first, reduced glutathione is an excellent biological free radical scavenger (Fantone & Ward, 1982) and second, it is the substrate for the reaction catalysed by glutathione peroxidase. Hence, in the absence of glutathione reductase, glutathione peroxidase has only limited protective capabilities. It is therefore important that these two enzymes be considered as a single protective system. The last point is illustrated in Fig. 1, which shows that the two enzymes, glutathione peroxidase and glutathione reductase, combine to cycle glutathione during the conversion of hydrogen peroxide to water. Figure 1 also shows the roles of superoxide dismutase and catalase. It must be clearly understood, however, that Fig. 1 does not pretend to be a definitive statement about the function of these enzymes, which are also extremely important in the removal of reactive species generated by the interactions of oxygen derived free radicals with cellular components. Experiments with the acetaldehyde-xanthine oxidase enzyme system suggest that hydrogen peroxide and the hydroxyl radical are not involved in killing adult N. brasiliensis in vitro . As catalase exerted no protective effect it seems safe to conclude that the
Free radicals and hexing
FIG. 1. The interactions of the four radical scavenging enzymes listed in Table 4. Abbreviations: SOD, superoxide dismutase; GP, glutathione peroxidase; CR, glutathione reductase; GSH, reduced glutathione; GSSG, oxidised glutathione.
active agent is not hydrogen peroxide. This also rules out involvement of the hydroxyl radical which is formed from hydrogen peroxide. The results obtained with 2,3_dihydroxybenzoic acid provide additional support for this view. 2,3-Dihydroxybenzoic acid is an iron-chefator (Graziano, Grady & Cerami, 1974) and therefore prevents the formation of the hydroxyl radical by Fenton’s reaction. Although it promoted the survival of N. brasiliensis it was not dependent on the compound’s iron-chelating abilities because 2,3_dihydroxybenzoic acid saturated with iron before incubation also enhanced survival. The chelator may thus act as a simple antioxidant through its loosely bound hydroxyl hydrogens and therefore cannot be regarded as a specific inhibitor of hydroxyl radical-mediated damage. The results in Tables 1 and 2 indicate that the superoxide radical is the most important mediator of free radical damage to N. brusiliensis, an idea which is backed up by the marked protective effect of superoxide dismutase. This result is perhaps slightly surprising, as it is generally accepted that the hydroxyl radical is the most reactive and therefore the most dangerous free radical in biological systems. However, free radical reactions which occur in vitro may not represent the sum of those that occur in vivo. Indirect evidence that the hydroxyl radicat may be implicated in the expulsion process is suggested by the fact that iron depletion significantly delayed expulsion of N. brusiliensb from the small intestine of the rat (Bolin, Davis, Cummins, Duncombe & Kelly, 1977). t-Butylhydroperoxide was apparently an effective generator of free radicals against N. brasiliensis in
vitro, as its effect was si~i~candy reduced by various antioxidants. The failure of superoxide dismutase to protect against t-butylhydroperoxide is to be expected as the superoxide radical is not produced by this compound. Rather, it probably generates peroxy and alkoxy radicals and possibly the hydroxyl radical. The protective effect of catalase is a strong argument in favour of peroxy and alkoxy radical generation though it does not rule out hydroxyl radical involvement. Indeed, the even greater protective effects of non-specific antioxidants such as tbiourea, mannitol, 2,5-dimethylfuran and 2,3_dihydroxybenzoic acid indicate that the hydroxyi radical may indeed be involved. The identity of the free radical which causes most damage to N. brasiliensik is, in the context of this paper, not so important as the demonstration that N. brasiliensis is more susceptible to free radical damage than N. dubius. Susceptibility to free radicals is probably related to levels of protective enzymes and may help to explain why N. br~il~e~~ infections undergo self-cure. Obviously an in vivo demonstration of free radical involvement in the self-cure phenomenon is essential, as the differences in protective enzyme levels in N. dubius and N. brasiliensis could be related to different degrees of dependence on aerobic metabolism by these parasites rather than a defence against host-generated free oxygen radicals. More active aerobic respirers might be expected to possess greater amounts of these enzymes. However, the relative oxygen uptake rates of the two parasites indicate that this is not the case. Nippostrongylus brasiliensis actually uses oxygen at a greater rate than N. dubius, which suggests that the protective enzymes
N. C. SMITHand C. BRYANT
in N. &&us are not solely a metabolic safeguard in respiration but a defence against host-generated free oxygen radicals. Alternatively, N. hrasiliensis may have more efficiently organised and cornpartitioned aerobic pathways than N. duhius, thus obviating the necessity for high levels of protective enzymes. Elucidation and comparison of fine details of the aerobic pathways of these two parasites is required. REFERENCES BASS D. A. & SZEJDA P. 1979. Eosinophils versus neutrophils in host defence. Killing of newborn larvae of Trichinella spiralis by human granulocytes in vitro. Journal of ClinicalInvestigation 64: 1415-1422. BERGMEYER H. U. 1974. Methods of Enzymatic Analysis, Vol. 1 (2nd edn). Verlag Chemie, Weinheim/Academic Press, New York and London. BOLIN T. D., DAVIS A. E., CUMMINS A. G., DUNCOMBE V. M. & KELLY J. D. 1977. The effect of iron and protein deficiency on the expulsion of Nippostrongylus brasiliensis from the small intestine of the rat. Guf 18: 182-l 86. BRADFORD M. 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72: 248-254. CHENG T. C. 1973. General Parasitology, pp. 637-639. Academic Press, New York. FANTONE J. C. &WARD P. A. 1982. Role of oxygen-derived free radicals and metabolites in leukocyte-dependent inflammatory reactions. American Journal of Pathology 107: 397-417.
GANSCHOW R. E. & SCHIMKLE R. J. 1969. Independent genetic control of the catalytic activity and the rate of degradation of catalase in mice. Journal of Biological Chemistry 244: 4649-4658. GRAZIANO J. H., GRADY R. W. & CERAMI A. 1974. The identification of 2,3-dihydroxybenzoic acid as a potentially useful iron-chelating drug. Journal ofPharmacology and-Experimental Therapeutics 190: 570-575. _. HOPKINS J. & TUDHOPE G. 1973. Glutathione oeroxidase in human red cells in health and disease. British Journal of Haematology 25: 563-575. KAZURA J. W. & MIZSHNICKS. R. 1984. Scavenger enzymes and resistance to oxygen-mediated damage in Trichinella spiralis. Molecular and Biochemical Purasitology 10: l10. MACKENZIEC. D., JUNCERYM., TAYLOR P. M. & OGILVIE B. M. 1981. The in vilro interaction of eosinophils, neutrophils, macrophages and mast-cells with nematode surfaces in the presence of complement and antibodies. Journal ofpathology 133: 161-175. MILLER H. R. P. 1971. Immune reactions in mucous membranes. III. The discharge of intestinal mast-cells during helminth expulsion in the rat. Laboratorylnvestigation 24: 348-354. OGILVIE B. M., HESKETH P.M. & ROSE M. E. 1978. Nippostrongylus brasiliensis: peripheral blood leucocyte response of rats with special reference to basophils. Experimental Parasitology 46: 20-30. SALIN M. L. & MCCORD J. M. 1974,Superoxidedismutases in polymorphonuclear leukocytes. Journal of Clinical Invesfigation 54: 1005-1009.