Human macrophage-mediated cytotoxicity of Schistosoma mansoni

Functional and structural features of the effector cells

B. J. COTTRELL1, C. PYE1, A. M. GLAUERT2 and A. E. BUTTERWORTH1

1Department of Pathology, University of Cambridge, Tennis Court Road, Cambridge CB2 IQP, UK2Strangeways Research iMboratory, Cambridge CBI 4RN, UK

Summary

Human monocytes were purified from peripheralblood and cultured in vitro on hydrophobic mem-branes. Such cells developed into mature tissue-type macrophages after approximately 1 week inculture. During this maturation period the macro-phages developed a potent cytotoxic mechanismwhereby they could kill the schistosomula of Schis-tosoma mansoni in standard in vitro cytotoxicityassays. Cytological and ultrastructural studies ofthe cells grown in vitro indicated that macrophagesdeveloped many of the classical histological andultrastructural features of 'activated' cells 'withruffled plasma membranes and significant in-creases in rough endoplasmic reticulum and Golgi

vesicles. Effective cytotoxicity depended upon con-tact of the effector cells and their parasite target.Further, experiments using metabolic inhibitorsindicated that cytotoxicity was dependent uponprotein synthesis. Initial results point to the macro-phage factor being distinct from some of the better-characterised macrophage secretory products suchas tumour necrosis factor, proteases and productsof oxygen metabolism.

Key words: monocyte, macrophage, cytotoxicity, Schistosomamansoni, in vitro.

Introduction

Macrophages grown in vitro can kill a range of targetsincluding tumour cells, microorganisms and tissue para-sites (James etal. 1982). In some instances, cytotoxicdestruction of the targets by the macrophages can occurin the absence of specific antibody to that target. Themolecular and ultrastructural basis of such cytotoxicity isbeing intensively investigated. To date many secretoryfactors have been identified including pre-eminentlyproteases, tumour necrosis factor (TNF), and productsof oxygen metabolism (Adams and Nathan, 1983; Drys-ddeetal. 1988).

The majority of the work has involved the use of mouseand rat peritoneal exudate cells or macrophage cell lines.However, attention has also focussed on the cytotoxic andfunctional potential of human macrophages (Musson,1983; Jungi and Hafner, 1986; Kreipe et al. 1988). Therelatively immature monocyte has little or no cytotoxicitytowards a range of tumour targets. However, as the cellsgrow into recognisable tissue macrophages many of theirsecretory functions increase in magnitude, and in parallelwith this there are large increases in cellular cytotoxicityof tumour targets (Hammerstr0m, 1979; Rinehart et al.1979; Andreesen et al. 1983, 1988).

Murine macrophages, in a manner analogous to tu-mour cytotoxicity, are also able to inflict damage upon the

Journal of Cell Science 94, 733-741 (1989)Printed in Great Britain @ The Company of Biologists Limited 1989

multicellular parasite Schistosoma mansoni and cells thathave been activated by interferon-gamma (IFN) kill invitro the schistosomula of this important human parasite(Bout et al. 1981; James et al. 19846). Further researchhas implicated the macrophage as an important potentialeffector cell in the expression of immunity in successfullyvaccinated mice (James et al. 1984a). The mechanism ofcytotoxicity in the mouse involves a macrophage se-cretory product that seems able to damage the internalorgans of the susceptible schistosomula (McLaren andJames, 1985; James and Glaven, 1987).

In contrast, human monocytes have demonstratedminimal ability to destroy schistosomula in vitro (Ellnerand Mahmoud, 1979). Work in our laboratories hasconfirmed that freshly isolated human monocytes possesslittle or no anti-schistosomula activity, but that this maybe increased by prior stimulation of the cells with IFN.More interestingly, unstimulated monocytes (with noadded IFN) could be grown in vitro into mature macro-phages that expressed very high levels of cytotoxicityagainst the parasite (Cottrell etal. 1989). Thus in thematuration of monocyte into macrophage the cellacquired a potent schistosomulicidal mechanism, whichwas fully expressed between 5 and 7 days in culture.

The present work therefore was undertaken to definemore, precisely the nature of cytotoxicity in our develop-ing macrophages. As a starting point we have looked for

733

cytotoxic mechanisms already described for the in vitromacrophage destruction of tumour targets (James et al.1982). Additionally, we considered such an approach mayprovide us with new insights into the changing interac-tions of the maturing effector cells with defined mediatorsof macrophage activation.

Materials and methods

The methods used in the present study have already beendescribed in detail elsewhere, so only a brief outline will begiven here (Cottrell et al. 1989).

Monocyte purification and cultureSamples of 100 ml of blood were obtained from healthy donors.These samples were then defibrinated and mixed with 12 ml ofdextran (6% dextran 500, Pharmacia; in Hanks' balanced saltsolution, HBSS), and 12ml of 2.7%'EDTAin HBSS. After30min the leucocyte-enriched supernatant was removed, and7 ml samples were spun over 3 ml of Nycodenz (Nycomed, UK)for 15min at 600 £ at 20°C (B0yum, 1983). Monocytes wereharvested from the interface, above the Nycodenz, and washedtwice with cold HBSS. Cells were counted with a Coultercounter, and cell viability was measured using the Trypan Blueexclusion method. Cytospin preparations were stained withGiemsa and for non-specific esterase (Koski etal. 1976). Themean number of viable monocytes from 100 ml of venous bloodwas21Xl06.

Cells were resuspended after two further washes with HBSS,at a concentration of SxlO5 cells per ml in RPMI with 10%heat-inactivated, pooled AB serum. These cells were thencultured on Petriperm, hydrophobic culture dishes (W. C.Heraeus, Hanau) in 5 % CO2, at 37°C (Andreesen et al. 1983).Cells were harvested at set intervals by chilling the plates for60min at 4°C, followed by vigorous pipetting. The macro-phages were washed twice with cold HBSS, counted and cellviability was assessed before use in the cytotoxicity assays.

Cytotoxicity assayTriplicate 0.S ml tubes (Eppendorf, F. R. Germany) containedSxlO5 cells; and 50 live schistosomula in 0.2ml of medium(Eagle's Minimal Essential Medium; 20mM-Hepes, sup-plemented with 100 units ml" of penicillin and lOOjUgml"1 ofstreptomycin, plus 10% heat-inactivated foetal calf serum;Flow). Inhibitors dissolved in the same medium were added togive the appropriate final concentration in a total volume of0.3 ml per tube. The cells and schistosomula (with or withoutinhibitors) were co-cultured for a period of 48 h at 37 °C in anatmosphere containing 5 % CC>2- Control cultures were set upseparately, either in plain medium or with inhibitor so that theeffects of inhibitors on effector cells and schistosomula could beevaluated (see Results and Table 2, below).

ReagentsHuman recombinant interferon-gamma (IFN); human recom-binant tumour necrosis factor (TNF); and a mouse monoclonalantibody to human TNF were all kind gifts from Dr G. R.Adolf (Ernst Boehringer-Institut fur Arzneimittel Forschung,Vienna, Austria). The inhibitors used in the study are listed inTable 1. These were obtained from Sigma (UK). Separateexperiments showed that at the concentrations used in theinhibitor experiments there was no significant toxic effect of anyof the inhibitors directly on the schistosomula (Table 2). Insome experiments macrophages were stimulated with a mixture

Table 1. The metabolic inhibitors used in the presentstudy

Inhibitor of: Concentration

Oxygen burstCatalase (bovine liver) (a,b,c)Cytochrome c (type III) (b)Sodium azide (d)Chlorpromazine (b)Dimethyl sulphoxide (e)

Lysosome functionHydrocortisone (b)Trypan Blue (b)Aurothioglucose (b)Ammonium chloride (f)Chloroquine (f)

Protein synthesisCycloheximide (f)Actinomycin D (f)

ProteasesTosyllysylchlormethylketone (f)Alpha-1 anti-trypsin (f)

MicrotubulesColchicine (f)

Iron-binding proteinsFerrous sulphate (g)

2X103 to 2X104 units ml"1

1 mgml"1

1 mM10 mM150 mM

100,1/gmr'1 mgml"1 mgml"1

10 mM1/JM-100//M

1-10 //g ml"1

1-10 ug ml" '

100 fiM100-500/<gml"'

1 mM

SO-100/iM

Also shown are references to the relevant papers (in parenthesis),(a) Wozencraft et al. (1984); (b) Malkin et al. (1987); (c) Adams andNathan (1983); (d) Kazura et al. (1981); (e) Rao et al. (1988);(f) James and Glaven (1987); (f) Drapier and Hibbs (1986).

Table 2. The effects of metabolic inhibitors on themacrophage cytotoxicity of schistosomula in vitro

Percentage cytotoxicity

Inhibitor

Normal mediumActinomycin D: l/(gml 1

- 1

- 1_1

No. expts

8837523

Cells

94±213±3«16±5*<37+10'46±8*'44±12'83±5

No cells

8±3* 6±5" 10±2•• 7±2' 15±6• 11±2

7±2

Cycloheximide: l/(gml10;(gml

Trypan BlueNH4C1

The results are expressed as the mean percentage of dead parasitesplus or minus the standard error of the mean.

*P<0.05; **P<0.001.

of calcium ionophore (A23187, 1 fa/i) and Escherichia coli LPS(0128: B12; lngml" 1 ) , both dissolved in normal medium(Sigma).

Schistosoma mansoni: life cycle and schistosomulapreparation

Details of the parasite and its routine laboratory maintenancehave been given elsewhere (Veith and Butterworth, 1983).Schistosomula were prepared by allowing infective cercariae topenetrate isolated rat skin in vitro, according to the method ofClegg and Smithers (1972).

Electron microscopyPellets of monocytes or macrophages were fixed in 2.5%

734 B. J. Cottrell et al.

glutaraldehyde in 0.09M-cacodylate buffer, pH7.2, containing3 mM-calcium chloride, for 30min at 20°C. The fixative wasthen removed and replaced with 0.1 M-cacodylate buffer,pH 7.2, with 3 mM-calcium chloride. The fixed pellets of cellswere stored at 4°C before further processing. The pellets werepost-fixed in 1 % osmium tetroxide in 0.1 M-cacodylate buffer,pH7.2, containing 3 mM-calcium chloride, for 1 h at roomtemperature, dehydrated in ethanol and embedded in Araldite.Thin sections (50-80 nm) were stained with lead citrate andexamined in a Philip's 201C microscope operating at 60 kV.

StatisticsIn the cytotoxicity experiments parasite attrition was scored bycounting the numbers of dead schistosomula (x), and thenumbers of live schistosomula ()>). In the tables and graphs,data are presented as percentage cytotoxicity:

- X 100.

100

80-

x+y

In these experiments simple Student's /-tests were performedon the log-transformed cytotoxicity data.

Results

Onset of anti-schistosomula activityPurified human monocytes were grown in vitro for 7days. On specified days cells were harvested and testedagainst live schistosomula in the standard cytotoxicitytest. The results, which are the pooled data from 14individual experiments, showed that freshly isolatedmonocytes (day 0) killed less than 10% of the schisto-somula targets. This did not achieve a significant differ-ence from the control level of kill with no cells in theculture (9 % ) . However, from day 2 onwards there was amarked increase in cytotoxicity of the parasites, so that byday 7 of culture the macrophages could kill 100% of theschistosomula (Fig. 1). Cells 5-7 days old were thereforevery effective cytotoxic macrophages and these weresubsequently used in all the experiments.

Initial experiments were also undertaken to assess theminimal numbers of macrophages required to kill eachschistosomulum. At a ratio of 10 effectors: target (log 4),cytotoxicity was 100 %. Even when this was reduced to alog ratio of 3.5, cytotoxicity exceeded 90% (Fig. 2).

100r

80

x 60oo§40

20

X

day 0 day 2 day 5 day 7Age of cells (days in culture)

Fig. 1. The onset of anti-schistosomula cytotoxicity withgrowth from the monocyte into the macrophage over a 7-dayperiod. The results are expressed as the mean cytotoxicity(from 14 experiments) plus or minus the standard error(vertical bars).

i1U 40

20

0 3 3.5 4log ratio

Fig. 2. The effect of effector: target ratio on the in vitrodestruction of the schistosomula. The ratios are expressed asthe log of the number of effector cells per singleschistosomulum.

From this we concluded that a ratio of 104 provided asuitable excess of cells in our standard assays.

Monocyte-macrophage morphologyExamination of thin sections in the electron microscopeshowed that the freshly isolated cells have the character-istic ultrastructure of human blood monocytes, withsmooth outlines, large indented nuclei, a limited Golgiregion, small rounded mitochondria and only a fewcisternae of the rough endoplasmic reticulum (Fig. 3).After incubation in vitro for as little as 2 days themonocytes developed into typical macrophages, withmany lamellipodia and surface blebs (Fig. 4). The mito-chondria were elongated and in some regions the roughendoplasmic reticulum enlarged to form extensive arraysof cisternae (Fig. 5). The Golgi region was also greatlyenlarged and contained arrays of smooth cisternae and avariety of vesicles and granules (Fig. 6). Thus the cellshave all the morphological characteristics of the maturetissue macrophage.

The findings of light microscopy confirmed this viewof the in vitro development of the monocytes into maturehistiocytic cells. The cells enlarged and, over a period of 7days displayed increasing amounts of granular cytoplasm.Also there were conspicuous ruffled plasma membranes;and enlarged complex nuclei, including multinucleateforms in some instances (Fig. 7).

Effects of metabolic inhibitors upon macrophagecytotoxicityGroups of inhibitors were used with known effects oncytotoxic macrophages, as listed in Materials andmethods (Table 1). Using cells that were 5-7 days old, anaverage of 94 % schistosomula were killed in the in vitrotests. However, both inhibitors of protein synthesis,namely cycloheximide and actinomycin D, causedmarked reductions in the level of parasite attrition.Actinomycin D was particularly effective in this regard,reducing cytotoxicity to a minimum of 13% (Table 2).Cell viability in the highest concentrations of actinomycinD and cycloheximide was 72% and 76%, respectively,compared to 81 % viability of control cells cultured inmedium alone.

Macrophage cytotoxicity of Schistosoma mansoni 735

Fig. 3. A freshly isolated monocyte has a large, indented nucleus (n), small rounded mitochondria (m), a few cisternae ofrough endoplasmic reticulum (er) and a small Golgi region (G). Electron micrograph. Bar, 1 fim.Fig. 4. After incubation for 2 days in vitm a monocyte has developed into the typical macrophage with surface lamellipodia. n,nucleus. Electron micrograph. Bar, l^im.

The only other inhibitor that successfully reducedcytotoxicity was Trypan Blue, which caused a reductionto 44%. Cell viability after 48 h was 83%. Althoughanother inhibitor of lysosomal function, ammonium

chloride, reduced cytotoxicity to 83 % this did notachieve statistical significance. Three other inhibitors oflysosomes, namely chloroquine, aurothioglucose and hy-drocortisone, had no effects on macrophage killing.

736 B. J. Cottrell et al.

Fig. 5. The cytoplasm of a mature macrophage contains elongated mitochondria (m) and numerous cisternae of theendoplasmic reticulum (er). n, nucleus; 1, lipid droplets. Electron micrograph. Bar, 1 j.im.

None of the oxygen-burst scavengers had any effectsupon cytotoxicity including cytochrome c, catalase,azide, chlorpromazine and dimethyl sulphoxide(DMSO) (all tested in two separate experiments). Twosubstances that interfere with trypsin function, tosyl-lysylchlormethylketone (TLCK) and alpha-1 anti-tryp-sin had no effect on the levels of macrophage killing.Colchicine, which inhibits microtubule polymerisation,similarly had no inhibitory action on the macrophages.Finally, excess iron in the form of FeSC>4 had no effect,indicating that the effector molecule did not bind iron.

The effects of TNFTNF has a well-documented cytotoxic action on anumber of tumour and microbial targets (Beutler andCerami, 1988). We investigated therefore the ability of

pure rTNF (recombinant TNF) to kill schistosomula invitro over 48 h. Further, TNF has been implicated as yetanother macrophage stimulator, so TNF was added tocultures of macrophages that were 1 day old. Such cellshave minimal cytotoxic effects, but do have the potentialto increase following stimulation with IFN, for example(Cottrell etal. 1989). In two experiments the back-ground cytotoxicity of schistosomula over 48 h (in me-dium only) was 6 ± 3 % . No concentration of TNF(2X 104, 2X 105 or 2X 106unitsml"1) either on its own oracting with additional cells caused a significant increase inpercentage cytotoxicity. For instance, at the highestTNF concentration there was 8 ± 5 % cytotoxicity withcells, and 14±5 % without cells.

In further experiments we added mouse monoclonalantibody to human rTNF to the cultures. A total of 104

neutralising units were added to each tube containing the

Macrophage cytotoxicity of Schistosoma mansoni 737

Fig. 6. Electron micrograph of a 2-day-old macrophage. The large Golgi region contains smooth cisternae (G) and a variety ofvesicles and granules, n, nucleus. Bar, 1 fim.

macrophages and schistosomula. In three experiments,macrophages (5 or 7 days old) killed 88±8 % of schistoso-mula. The addition of the monoclonal antibody to thecultures had no effects on macrophage killing of theparasites, the mean cytotoxicity being 86±8%. Withoutcells cytotoxicity was just 2%.

Arginine: and effects of arginaseArginase has been shown to be cytotoxic for tumours andhas been implicated as a potential cytotoxic factor forschistosomes (Olds etal. 1980). Also arginine, in someexperimental tumoricidal systems, is a prerequisite forthe synthesis by the macrophage of cytotoxic factor thatinhibits tumour mitochondrial respiration (Hibbs et al.1987). To look at the effects of both these substances weconstructed the following experiment. Five-day-old mac-rophages were used against schistosomula in either theconventional complete MEM medium, medium depletedof all amino acids, or medium without amino acids butwith additional arginine. We also tested the effects ofarginase (5 units mP 1 ) in complete medium. The resultsof two experiments are summarised in Table 3. They

Table 3. The effects of eliminating amino acids (-AA)from the medium, and the presence of arginase on the

macrophage cytotoxicity

CellsNo cells

MEM**

10013±1

Culture

MEM-AA

10014±8

medium

MEM-AA

-Harginine

10015±2

MEM+arginase

10016+3

The results are expressed as the mean percentage cytotoxicity plusor minus the standard deviation.

**MEM is complete medium with all amino acids (AA).

showed that no additional amino acids were required inthe medium for the expression of cytotoxicity, and thatschistosomula could survive in the depleted media ifmacrophages were absent. Further, arginase on its ownhad no cytotoxic effects on the schistosomula.

738 B. Jf. Cottrell et al.

Fig. 7. Giemsa-stained cytospin preparations of freshly isolated monocytes (A); 2-day-old macrophages (B); and 7-day-oldmature macrophages (C and D). All preparations at the same magnification, photographed under oil immersion. Note theenlargement of the cells with the increasing complexity of the cell nuclei; the development of abundant granular cytoplasm; andin (D) the 'ruffled' plasma membrane. Also note the presence of lipid droplets within the cytoplasm of the cells. XlOOO.

The requirement for macrophage-parasite contactTo see whether physical contact of effector and target wasa necessary element in the successful killing of theparasite, we utilised the standard Millipore chambers(0.45 (im pore size) to separate the components of ourcytotoxic assay. Thus 2xlO6 macrophages were culturedin the chambers with 100 schistosomula; or cells werecultured alone with schistosomula in the 12 mm diameterwell that housed the Millipore chamber. All combi-nations of cells and parasites were cultured in a totalvolume of 0.5 ml of medium. The results showed thatseparation of the effector cells and the parasites byMillipore filter reduced cytotoxicity to 11 ±4 % comparedto 100% when macrophages had uninterrupted access tothe schistosomula (no filter).

Activity of macrophage supernatantsTo investigate the possibility that macrophages werecapable of secreting factors toxic for schistosomula welooked at the anti-schistosomula activity of media con-ditioned by our 7-day-old cells. Initial results indicated,however, that unstimulated cells produced little or noactivity in the supernatants. We attempted therefore toincrease the secretory potential of the macrophages bystandard regimes of stimulation involving IFN followedby E. coli LPS and calcium ionophore (Zacharchuk et al.1983). Accordingly, 7-day-old cells were cultured over-night with IFN (104 units ml~ ), followed by washingand a second 30-min stimulation with a combination ofLPS and ionophore. The cells were washed once again,and then cultured in MEM for 2 h at a concentration of

Macrophage cytotoxicity of Schistosoma mansoni 739

106 cells ml . Supernatants from these macrophageswere then tested against schistosomula over 48 h. Innormal, unconditioned medium there was a backgroundkill of 13±4%. The results showed that stimulation ofthe cells with IFN alone did not cause a statisticallysignificant increase in the cytotoxic activity of the super-natants (19±2%). However, macrophages treated withLPS and ionophore did show statistically significantincreases in the secretion of anti-schistosomula factor(s)into the conditioned medium. Further, this seemed to beindependent of whether the cells were given an initialexposure to the IFN (cytotoxicity 31±5%); or not(37±4%), before stimulation with LPS and ionophore(both results significant; P<0.05).

Discussion

Macrophages are able to destroy a range of microorgan-isms, and tumour cell targets (Andrew etal. 1985;Drysdale et al. 1988). Crucial to an analysis of cytotox-icity is an understanding of the mechanisms of macro-phage activation and maturation (Adams and Hamilton,1984). The better documented mechanisms of tumourcell killing include a range of macrophage secretoryproducts such as cytotoxic proteases (Adams andNathan, 1983); TNF (Matthews, 1981); lysosomal en-zymes (Leoni etal. 1985); oxygen metabolites (Nathanetal. 1983); and arginase (Currie and Basham, 1978).

Macrophage-mediated destruction of the schistoso-mula bears many similarities to tumour cytotoxicity in invitro systems that utilise mouse peritoneal exudate cells(James etal. 1982; McLaren and James, 1985). How-ever, the role of the human monocyte and tissue macro-phage is less clear. We present evidence here to show thatthe acquisition of cytotoxicity by monocytes is dependentupon a maturational event. A parallel developmentalchange is seen when monocytes and in vitro culturedhuman macrophages are tested against tumour cell tar-gets. Monocytes are relatively ineffectual, but levels oftumour cytotoxicity increase as the cells develop intotypical tissue macrophages over periods of 7 to 14 days(Hammerstr0m, 1979; Rinehart et al. 1979; Andreesenetal. 1983). As the cells develop and increase theirprotein content there are significant increases in lyso-somal enzymes (Musson, 1983); whilst other functionssuch as the generation of oxygen metabolites are down-regulated (Nakagawara etal. 1981). The relative contri-bution of these cytotoxic mechanisms will probablytherefore be a function of the developmental status anddegree of activation of the human macrophage (Andree-sen et al. 1988; Kreipe et al. 1988).

Light and electron microscopy of the growing mono-cyte confirms the typical development into a tissue-type,mature macrophage. There is a characteristic large in-crease of rough endoplasmic reticulum, and associatedGolgi apparatus, indicating a high level of proteinsynthesis, and potential secretory activity (Carr, 1980).In this context, it is an interesting finding that inhibitorsof protein synthesis are the most effective agents in theabrogation of the macrophage cytotoxicity. Similar re-

sults in mouse macrophages have also been obtained byother workers (James and Glaven, 1987). Thus, theresults from our metabolic and morphological studiesimplicate active protein synthesis as one componentnecessary for effective macrophage killing of the parasitein vitro.

Stimulated macrophages treated with agents known todisrupt membranes, release significant cytotoxic activityinto the supernatant. This may be the consequence of lowlevels of secretion in 'unstimulated' macrophages, or itmight indicate that the effector molecules are membrane-bound. Certainly, for the unstimulated macrophage (i.e.not treated with ionophore and LPS) intimate contactwith the schistosomula targets is a necessary prerequisitefor cytotoxicity. Separation of effector cells and targets bya Millipore filter, which allows for the passage of mediumbut not cells, completely abolishes effective parasitekilling. We may speculate that such contact promoteslocally high concentrations of secreted molecules, or evenpromotes the interchange of membrane-bound materialfrom the effector cells to the parasite surface.

We present evidence here that some of the macrophagesecretory products outlined above are probably notimportant in the destruction of the schistosomula. Forinstance, it is unlikey that TNF is the effector molecule,as monoclonal antibody to rTNF is unable to block thenormal macrophage killing; and rTNF has no effect onthe in vitro cultured schistosomula. Similarly, the inhi-tion experiments using oxygen radical scavengers indicatethat the products of oxygen-burst metabolism are unliklycandidates for the effector molecules in our system. Thisresult reinforces earlier work in the mouse that shows thatnon-oxidative metabolic pathways are operational in themacrophage destruction of the parasite (James and Gla-ven, 1987; Malkin etal. 1987). Of the remaining non-oxidative mechanisms, we find that trypsin is anotherunlikely candidate, as two agents known to block itsactivity have no effect in our cytotoxicity assays.

Ultrastructural studies of the interaction of mousemacrophages and schistosomula show that damage isdirected at mitochondria and smooth muscle (McLarenand James, 1985). It is intriguing therefore to find thatactivated mouse macrophages may kill tumour targets byinhibiting mitochondrial respiration (Kilbourn et al.1984). More recent experimental evidence has clearlyshown that this mechanism of tumour cell damage isdependent upon the metabolism of arginine in themacrophages (Hibbs et al. 1987). Interesting as theserecent studies are, there is no evidence from our systemthat cytotoxicity is the result of an arginine-dependentmetabolic pathway.

Further work is being carried out in our laboratory todefine the mechanism of cytotoxicity of the schistoso-mula. In this context, the use of defined populations ofinactive monocytes and maturing cytotoxic macrophagesis proving useful.

This work was supported by the Wellcome Trust, London(B.J.C.), and the Sir Halley Stewart Trust. We thank Mr R.Parker for his excellent technical assistance in the preparation of

740 B. J. Cottrell et al.

the electron micrographs. We also thank Ms J. Jackson for herassistance throughout the course of this project.

References

ADAMS, D. O. AND HAMILTON, T. A. (1984). The cell biology ofmacrophage activation. A. Rev. Immim. 2, 283-348.

ADAMS, D. O. AND NATHAN, C. F. (1983). Molecular mechanisms intumour-cell killing by activated macrophages. Immunol. Today 4,166-170.

ANDREESEN, R., GADD, S., BRUGGER, W., LOHR, G. W. ANDATKINS, R. C. (1988). Activation of human monocyte-derivedmacrophages cultured on Teflon: response to interferon-gammaduring terminal maturation in vitro, lmmunobiology 177, 186-198.

ANDREESEN, R., OSTERHOLZ, J., BROSS, K. J., SCHULZ, A.,LUCKENBACH, G. A. AND LOHR, G. W. (1983). Cytotoxic effectorcell function at different stages of human monocyte-macrophagematuration. Cancer Res. 43, 5931-5936.

ANDREW, P. W., JACKETT, P. S. AND LOWDRIE, D. B. (1985).Killing and degradation of microorganisms by macrophages. InMononudear Phagocytes: Physiology and Pathology (ed. R. T.Dean and W. Jessup), pp. 311-335. Elsevier Science PublishersB.V. (Biomedical Division), Amsterdam.

BEUTLER, B. AND CERAMI. A. (1988). The common mediator ofshick, cachexia and tumour necrosis. In Advances in Immunology,vol. 42 (ed. F. R. Dixon), pp. 213-231. Academic Press, New-York, London.

BOUT, D. T., JOSEPH, M., DAVID, J. R. & CAPRON, A. R. (1981). Invitro killing of S. mansoni schistosomula by lymphokine-activatedmouse macrophages. J. Immun. 127, 1-5.

B0YUM, A. (1983). Isolation of human blood monocytes withNycodenz, a new non-ionic iodinated gradient medium. Scand.J. Immun. 17, 429-436.

CARR, I. (1980). In The Retioculoendothelial System. AComprehensive Treatise, vol. 1 (ed. I. Carr and W. T. Daems),pp. 671-703. New York, London: Plenum Press.

CLEGG, J. A. AND SMITHERS, S. R. (1972). The effects of immunerhesus monkey serum on schistosomula of Schistosoma mansoniduring cultivation in vitro. Int. J. Parasit. 2, 79-98.

COTTRELL, B. J., PYE, C. AND BUTTERWORTH, A. E. (1989).Cytotoxic effects in vitiv of human monocytes and macrophages onthe schistosomula of Schistosoma mansoni. Parasit. Immun. 11,91-104.

CURRIE, G. A. AND BASHAM, C. (1978). Differential argininedependence and the selective cytotoxic effects of activatedmacrophages for malignant cells in vitro. Br.J. Cancer. 38,653-659.

DRAPIER, J. AND HIBBS, J. B. JR (1986). Murine cytotoxic activatedmacrophages inhibit aconitase in tumour cells. J. din. Invest. 78,790-797.

DRYSDALE, B.-E., AGARWAL, S. AND SHIN, H. S. (1988).Macrophage-mediated tumoricidal activity: mechanisms ofactivation and cytotoxicity. Prog. Allergy 40, 111-161.

ELLNER, J. J. AND MAHMOUD, A. A. F. (1979). Killing ofschistosomula of Schistosoma mansoni by normal humanmonocytes. J . Immun. 123, 949-951.

HAMMERSTR0M, J. (1979). Human macrophage differentiation in vivoand in vitro. Acta path, microbiol. scand. sect. C 87, 113-120.

HIBBS, J. B. JR, TAINTOR, R. R. AND VAVRIN, Z. (1987).Macrophage cytotoxicity: role for L-arginine deiminase and iminonitrogen oxidation to nitrite. Science 235, 473-476.

JAMES, S. L., CORREA-OLIVEIRA, R. AND LEONARD, E. J. (1984a).Defective vaccine-induced immunity to Schistosoma mansoni in Pstrain mice. II. Analysis of cellular responses..7. Immun. 133,1587-1593.

JAMES, S. L. AND GLAVEN, J. A. (1987). Effects of inhibitors oftumoricidal activity upon schistosomulum killing by activatedmacrophages. Infect. Immun. 55, 3174-3180.

JAMES, S. L., LEONARD, E. J. AND MELTZER, M. S. (1982).

Macrophages as effector cells of protective immunity in murineschistosomiasis. IV. Coincident induction of macrophage activationfor extracellular killing of schistosomula and tumour cells. Cell.Immun. 74, 86-96.

JAMES, S. L., NATOVITZ, P. C , FARRAR, W. L. AND LEONARD, E. J.(19846). Macrophages as effector cells of protective immunity in

murine schistosomiasis: macrophage activation in mice vaccinatedwith radiation-attentuated cercariae. Infect. Immun. 44, 569-575.

JUNGI, T. W. AND HAFNER, S. (1986). Quantitative assessment of Fcreceptor expression and function during in vitro differentiation ofhuman monocytes to macrophages. Immunology 58, 131-137.

KAZURA, J. W., FANNING, M. M., BULMER, J. L. AND MAHMOUD,A. A. F. (1981). Role of cell-generated hydrogen peroxide ingranulocyte-mediated killing of schistosomula of Schistosomamansoni in vitro. J. din. Invest. 67, 93-102.

KlLBOURN, R. G. , KLOSTERGAARD, J. AND LOPEZ-BERESTEIN, G.(1984). Activated macrophages secrete a soluble factor that inhibitsmitochondrial respiration of tumour cells. J. Immun. 133,2577-2581.

KOSKI, I. R., POPLACK, D. G. AND BLAESE, R. M. (1976). In In

Vitro Methods in Cell-mediated and Tumour Immunity (ed. B. R.Bloom and J. R. David), pp. 359-361. New York, San Francisco,London: Academic Press.

KREIPE, H., RADZUN, H. J., RUDOLPH, R., BARTH, J., HANSMANN,

M. J., HEIDORN, K. AND PARWARESCH, M. R. (1988).Multinucleated giant cells generated in vitro. Terminallydifferentiated macrophages with down-regulated c-fms expression.Am. J. Path. 130,232-243.

LEONI, P., DEAN, R. T. AND JESSUP, W. (1985). Secretion of

hydrolases by mononuclear phagocytes. In MononuclearPhagocytes: Physiology and Pathology (ed. R. T. Dean and W.Jessup), pp. 181-202. Elsevier Science Publishers B.V.(Biomedical Division), Amsterdam.

MALKIN, R., FLESCHER, E., LENGY, J. AND KEISARI, Y. (1987). Onthe interactions between macrophages and developmental stages ofSchistosoma mansoni: the cytotoxic mechanisms involved inmacrophage-mediated killing of schistosomula in vitro.lmmunobiology 176, 63-72.

MATTHEWS, N. (1981). Production of an anti-tumour cytotoxin byhuman monocytes. Immunology 44, 135-142.

MCLAREN, D. J. AND JAMES, S. L. (1985). Ultrastructural studies ofthe killing of schistosomula of Schistosoma mansoni by activatedmacrophages in vitro. Parasit. Immun. 7, 315-331.

MUSSON, R. A. (1983). Human serum induces maturation of humanmonocytes in vitro. Am. J. Path. I l l , 331-340.

NAKAGAWARA, A., NATHAN, C. F. AND COHN, Z. A. (1981).Hydrogen peroxide metabolism in human monocytes duringdifferentiation in vitro. J. din. Invest. 68, 1243-1252.

NATHAN, C. F., MURRAY, H. W., WIEBE, M. E. AND RUBIN, B. Y.

(1983). Identification of interferon-y as the lymphokine thatactivates human macrophage oxidative metabolism andantimicrobial activity, j . exp. Med. 158, 670-689.

OLDS, G. R., ELLNER, J. J., KEARSE, L. A. JR, KAZURA, J. W. ANDMAHMOUD, A. A. F. (1980). Role of arginase in killing ofschistosomula of Schistosoma mansoni. J. exp. Med. 151,1557-1562.

RAO, P. S., LUBER, J. M. JR, MILLNOWICZ, J., LALEZARI, P. ANDMUELLER, H. S. (1988). Specificity of oxygen radical scavengersand assessment of free radical scavenger efficiency using luminolenhanced chemiluminescence. Biochem. biophvs. Res. Commun.150, 39-44.

RlNEHART, J. J., VESSELLA, R., LANGE, P., KAPLAN, M. E. ANDGORMUS, B. J. (1979). Characterization and comparison of humanmonocyte- and macrophage-induced tumour cell cytotoxicity.J. Lab. din. Med. 93, 361-369.

VEITH, M. C. AND BUTTERWORTH, A. E. (1983). Enhancement ofhuman eosinophil-mediated killing of Schistosoma mansoni larvaeby mononuclear cell products in vitro. J. exp. Med. 157,1828-1843.

WOZENCRAFT, A. O., DOCKRELL, H. M., TAVERNE, J., TARGETT, G.A. T. AND PLAYFAIR, J. H. L. (1984). Killing of human malariaparasites by macrophage secretory products. Infect. Immun. 43,664-669.

ZACHARCHUK, C. M., DRYSDALE, B.-E., MAYER, M. M. AND SHIN,H. S. (1983). Macrophage-mediated cytotoxicity: Role of a solublemacrophage cytotoxic factor similar to lymphotoxin and tumournecrosis factor. Proc. natn. Acad. Sci. U.S.A. 80, 6341-6345.

{Received 17 March 1989 - Accepted, in revised form, S September1989)

Macrophage cytotoxicity of Schistosoma mansoni 741