Anionic site behavior in Leishmania and its role in the parasite-macrophage

interaction

E. M. B. SARAIVA1'2, M. A. VANNIER-SANTOS1, F. C. SILVA-FILHO1 and W. DE SOUZA1'*

' Laboratdrio de Ullraestnitura Celular, Institute de Biofisica Carlos Chagas Filho and zDepartamento de Imunologia, Instituto de Microbiologia,Universidade Federal do Rio de Janeiro, Ilha do Fundao, 21941, Rio de Janeiro, Brasil

* Author for correspondence

Summary

The behavior of cationized ferritin (CF) bindingsites on the surface of Leishmania mexicana ama-zonensis (amastigotes, infective and non-infectivepromastigotes) and their participation in the inter-action with macrophages were evaluated. Glutaral-dehyde-fixed parasites treated with CF present auniform labelling over the whole cell surface. How-ever, living parasites displayed CF patches andcaps. Capping was usually seen towards the an-terior (flagellated) portion of the cells, where shed-ding phenomena took place. These processes wereinhibited by sodium azide but not by low tempera-ture (4°C). CF treatment of non-infective promasti-gotes led to an increase in their uptake by macro-phages, whereas the uptake of amastigotes orinfective promastigotes was not significantlyaltered. The effect of CF on the parasite surfacecharge was analyzed by whole-cell microelectro-phoresis. The mean electrophoretic mobility (EPM)

of non-infective promastigotes was decreased by26%, while once again the other parasite forms'were not significantly affected. Transmission elec-tron microscopy of mouse peritoneal macrophagecultures, fixed after interaction with CF-labelledparasites, revealed that both amastigotes and infec-tive promastigotes quickly removed bound CF.Therefore CF was seen neither in parasite-macro-phage attachment areas nor in parasitophorousvacuoles. On the contrary, non-infective promasti-gote-macrophage attachment areas were remark-ably large and preferentially comprised CF-labelled membranes. These results strongly suggestan important participation of cell surface anionicsites in the L. mexicana amazanensis-macrophageinteraction.

Key words: Leishmania-macrophage interaction, anionicsites, cationized ferritin.

Introduction

Parasites of the genus Leishmania are of special interest,since they infect cells of the mononuclear-phagocyticsystem. It is through the contact between the two cellsurfaces that this crucial step in the parasite's life cycletakes place. Promastigotes inoculated by the invertebratehost infect vertebrate phagocytes, in which they aretransformed and reproduce as amastigotes. The mechan-isms underlying such phagocyte specificity are still poorlyunderstood, although different surface receptor-ligandsystems, involving fibronectin, complement and man-nose 6-phosphate receptors, have been shown to partici-pate in this recognition process (Wyler et al. 1985;Russell &Wilhelm, 1986; Russell & Wright, 1988; Wilson& Pearson, 1986; Mosser et al. 1987; Saraiva et al. 1987;Puentes et al. 1988; Wozencroft et al. 1986). Alterna-tively, parasite-macrophage interaction may be studiedas a physical surface phenomenon. It is well established

Journal of Cell Science 93, 481-489 (1989)Printed in Great Britain © The Company of Biologists Limited 1989

that most cell types have a net negative surface charge,and that cell surface properties can participate in cellularphenomena such as recognition, adhesion and endo-cytosis (reviewed by Van Oss, 1978). In the case ofLeishmania, macromolecules such as the lipophosphogly-can (LPG) may be relevant because of its highly anionicnature (reviewed by Turco, 1988). The net negativesurface charge of macrophages and Leishmania has beenanalyzed by cell electrophoresis (Pimenta & De Souza,1983; Silva-Filho et al. 1987; Santos & De Souza, 1983)and by labeling with cationic probes (Papadimitriou,1982; Pimenta & De Souza, 1983; Santos & De Souza,1983), but no single work to date has attempted tocorrelate these data with the interaction process. Polyca-tionic substances such as polylysine (Capo et al. 1981;Matsui et al. 1983), lysozyme (Pruzanski & Saito, 1978)and cationized ferritin (Meirelles et al. 1984) enhanceendocytosis. Cationized ferritin has the advantage ofbeing readily identified on electron micrographs, because

481

of its electron density and particulate appearance. Incontrast to other cationic substances (Werk et al. 1984)CF can be used at physiological pH (Danon et al. 1972),without interfering with cell viability.

Among trypanosomatids, cationized ferritin binding toT. cruzi (De Souza et al. 1977; Souto-Padron & DeSouza, 1986) and to Leishmania (Pimenta & De Souza,1983; Ayesta et al. 1985) has been shown in aldehyde-fixed cells, and so the dynamics of this reaction and itspossible biological implications remain unknown. Thefate of cationized ferritin-binding sites has been studiedin T. cruzi and in Leishmama-macrophagz interactions,but mostly by focusing the formation of endocyticvacuoles (Meirelles et al. 1984; Pimenta & De Souza,1988).

In this paper we report the behavior of cationizedferritin-binding sites on the surface of living Leishmaniamexicana amazonensis (amastigotes, infective and non-infective promastigotes) and their participation in theinteraction with macrophages.

Materials and methods

ParasitesThe Josefa strain of Leishmania mexicana amazonensis ob-tained from Dr C. A. Cuba Cuba (Universidade de Brasilia,Brasil) in 1981 has been maintained in our laboratory sincethen, in culture and by hamster footpad inoculation. Promasti-gotes cultured for long periods in axenic medium are considerednon-infective (Saraiva et al. 1986) because they are unable toproduce visible lesions when inoculated into hamsters or tosurvive within macrophages maintained in vitro. The promasti-gotes used for the experiments described here had undergonemore than 450 passages in vitro. Freshly transformed infectivepromastigotes were obtained by culturing amastigotes fromhamster lesions for up to five sub-cultures. Both were main-tained at 26°C in a medium consisting of a brain-heart infusionsupplemented with 10% fetal calf serum and a 0-2% hemoglo-bin solution, and harvested for the experiments during thestationary phase of growth. Amastigote purification from ham-ster footpad lesions has been described in detail (Saraiva et al.1983).

MacrophagesPeritoneal macrophages from normal 6- to 8-week-old Swissmice were collected in Hank's balanced solution and plated onglass coverslips in Falcon 24-well tissue-culture plates (BectonDickinson Labware, New Jersey, USA) as described by Saraivaet al. (1987). The cells were allowed to adhere for 30-40 min at37°C in a 5 % CO2 atmosphere, after which time the non-adherent cells were removed and new culture medium (Eagle'splus 10% fetal calf serum) was added. The cells were thenincubated overnight under the same conditions as above beforethe interaction experiments.

Cationized ferritin treatmentBefore interaction, either the macrophages or the parasites wereincubated at 4°C for 30 min in the presence of cationized ferritinat 100jUgml~ or 10 jig ml" (subagglutinating concentration),respectively. After washing twice with Hank's solution at 4°Ccells and parasites were allowed to interact for 1 h at 37°C. Theparasites were also fixed in 2 5 % glutaraldehyde in 0 1 M -phosphate buffer, pH7-2, for the electrophoretic mobility

determination or processed for transmission electron mi-croscopy. For the capping experiments, promastigotes sus-pended in PBS (lOmM-phosphate buffer; 150mM-NaCl),pH7-2, were incubated with cationized ferritin (10;/gml~ ) for30 min at 4°C. The parasites were subsequently washed in PBSby centrifugation at 1000g for 10 min at 4°C. After that, onesample was immediately fixed, the others were incubated at26°C and portions were removed at 10, 20 and 30 min, and fixedas described below.

Parasite--macrophage interactionInteraction experiments were performed as described (Saraivaet al. 1987). Briefly, washed parasites and macrophages wereleft in contact for 1 h at 37°C in a parasite: macrophage ratio of50: 1 in Eagle's medium (Sigma Chem. Co., MO, USA). Thecells were then rinsed with Ringer's solution, fixed in Bouin andstained with Giemsa. The percentage of infected macrophageswas determined by randomly counting a minimum of 600 cellsin each of the triplicate preparations. Experiments were re-peated at least three times. For transmission electron mi-croscopy, the interactions were carried out in Descarplastculture flasks (Rio de Janeiro, Brasil) in a parasite: macrophageratio of 10:1. After 30 min of interaction the cultures were fixedin 4% paraformaldehyde, 1 % glutaraldehyde and 1 mM-CaCl2in O'l M-cacodylate buffer, washed overnight in the same bufferand, after removal with a rubber policeman, they were post-fixed for 1 h in a solution containing 1 % osmium tetroxide,0-8% potassium ferricyanide, 1 niM-CaO2 in 0-1 M-cacodylatebuffer, washed in the same buffer, dehydrated in acetone andembedded in Epon.

Cell microelectrophoresisThe electrophoretic mobility (EPM) was determined using aZeiss cytopherometer, by timing the passage of cells through acalibrated graticule, when a current of 6 mA and a gradient of5-5 Vcm"1 were applied to the electrophoresis chamber, asdescribed (Silva-Filho et al. 1987). Briefly, fixed parasites (in2-5% glutaraldehyde in 0-1 M-phosphate buffer, pH7-2) werecarefully washed and suspended in a NaCl solution (ionicstrength 0-145moldm~3, pH7-2) for the EPM determinations.Measurements of living cells at 4°C were not significantlydifferent from those obtained at 25 °C with glutaraldehyde-fixedcells (Pimenta & De Souza, 1983). The EPM was determinedusing the equation: EPM = d/t X D/V, where d is the distance(usually 16,um) the cell is conveyed during measurements, / isthe time (s) required for the cell to move along d, D is thedistance (18 cm) between the two electrodes and V is thepotential applied to the electrodes. For each experiment theEPM of at least 50 cells was timed in alternating directions, toavoid electrode polarization. Statistical analysis was done usingStudent's Mest. We considered significant those differencesabove 10% (P^0-05).

Results

Cationized ferritin binding to the Leishmania surfaceGlutaraldehyde-fixed parasites treated with cationizedferritin (CF) particles present uniform labeling over thewhole cell surfaces (Fig. 1). On the contrary, livingparasites preincubated at 4°C with CF, shifted to 25 °Cand fixed after 10-30 min, showed a non-uniform, patchydistribution of the particles (Fig. 2). This CF rearrange-ment was inhibited by azide but not by low temperature(4°C). Capping of the surface CF particles was seen at

482 E. M. B. Saraiva et al.

^ ^ ^ * V

4 jFig. 1. Glutaraldehyde-fixed promastigote of L. mexicana amazonensis treated with lOjUgml"1 cationized ferritin (CF).A uniform CF distribution is observed on both flagellar and cell body membranes./, flagellum; X25 000.Fig. 2. Promastigote preincubated with 10,ugml~ CF and kept at room temperature for lOmin before fixation. Note theuneven, patchy CF distribution at the anterior end of the cell (arrowheads)./, flagellum; X25 000.Fig. 3. CF patching (open arrows) and shedding (filled arrows) from the flagellar basis./, flagellum; X55000.Fig. 4. CF shedding in the presence of 3 mM-azide (see text). Membrane units (arrowheads) can be observed in association withliberated CF particles./, flagellum; X34000.

both poles of the cells, but far more frequently towardsthe flagellar, anterior end of the parasite. Membrane-bound CF particles were eventually released from flagel-lar membranes (Fig. 3) but CF-associated membraneunits were hardly seen (unless the preparation angle wasaltered using a goniometer), possibly because particleaggregation caused the twisting of detached membranes.Low azide concentrations (3 mM) also made membraneunits more easily observed (Fig. 4), probably because ofthe reduced amount of bound CF on shedding mem-branes.

Cationized ferritin-labeled free membranes were oftenfound in our preparations, especially with infectivepromastigotes, which appeared to remove bound CFparticles much more rapidly than non-infective ones.After 30 min at room temperature, most infective cellswere almost completely depleted of CF, whereas the non-infective cells still presented large amounts of particles.

Effects of cationized ferritin on theLeishmania—macrophage interactionCationized ferritin (CF) had different effects upon the

interaction 483

ParasiteMacrophage

(S.E.M.)

% of infected cells

8

8-96(1-24)

2-51(0-24)

1-69(0-11)

7-43(1-11)

2-70(0-46)

1-65(0-11)

6-57(0-81)

2-51

(0-60)1-64

(0-19)

v/m

V////AM

H

d

H

i

Ji

J

_

Control

CF-treatedLeishmania

CF-treatedmacrophages

Fig. 5. Effect of cationized ferritin (CF) on the L. mexicanaamazonensis (amastigotes, infective and non-infectivepromastigotes) macrophage interaction. Parasites ormacrophages previously treated with CF (10 or 100 f.ig nil" ,respectively) were allowed to interact for 1 h, fixed andGiemsa stained. Control of parasite-macrophage interaction(CF untreated): (D) amastigote; (@) infective promastigote;(M) non-infective promastigote. CF-treated parasites-macrophages: (D) amastigotes; ( S ) infective promastigotes;(M) non-infective promastigotes. Parasites-CF-treatedmacrophages: ( • ) amastigotes; (H) infective promastigote;(M) non-infective promastigote.

interaction of macrophages with amastigotes, and infec-tive and non-infective promastigotes of L. mexicanaamazonensis (Fig. 5). Treatment of the macrophageswith CF before the interaction significantly increased theuptake of non-infective promastigotes, mainly due to ahigher percentage of infected macrophages. However,the same treatment did not interfere with the uptake ofamastigotes and infective promastigotes. It is interestingto note that the mean number of parasites per infectedcell, as compared with controls, was not altered for bothpromastigotes. In amastigotes the same index decreasedby 27 % but the percentage of infected macrophages wasnot affected by the CF treatment of phagocytes.

In the same way, only the non-infective promastigoteincubation with CF increased its uptake by macrophages,again by producing a higher percentage of infected cells.At the light microscopy level it is not always possible todistinguish between attachment and actual uptake, butultrastructural observations of many thin sections suggestthat both processes were remarkably enhanced in CF-treated non-infective cells. Preincubation with CF ofinfective promastigotes and amastigotes did not signifi-cantly affect their interaction with macrophages.

Transmission electron microscopy of macrophage cul-tures fixed after interaction with CF-labelled parasites

Figs 6-12. L. mexicana amazonensis promastigotes weretreated with 10f«gml~ CF, at 4CC for 30min, beforeinteraction with mouse peritoneal macrophages.Fig. 6. Three non-infective promastigotes attached to amacrophage, showing CF in the attachment areas. X 12 500.Fig. 7. Parasite bound to two macrophages. CF can be seenonly in attachment areas, in, macrophage; p, parasite.X14 000.Fig. 8. Non-infective promastigote during ingestion bymacrophage. A macrophage pseudopod is observed covering aCF-labelled membrane area. X22500. Inset: macrophagepseudopods preferentially bind to CF-labeled areas. X45 500.Fig. 9. CF-labeled flagellum (/) attached to a macrophage.x 35 000.Fig. 10. CF removal (arrow) from bound parasite (see text).X 34 000.Fig. 11. Macrophage showing a vacuole (v) with endocytosedCF-labeled membrane. Extracellular CF-labeled membranesare observed attached to the macrophage surface(arrowheads). X22000.Fig. 12. Larger vacuole (v) showing several CF-labeledvesicles, x22 000.

showed different labelling patterns for infective and non-infective cells. Both infective promastigotes and amasti-gotes quickly removed bound CF, whereas non-infectivecells presented CF particles during attachment(Figs 6-10) and even after endocytosis (Fig. 13).

Non-infective promastigotes attachment areas wereremarkably large (Figs 6 and 8) and preferentially me-diated by CF-labelled membranes. The CF layer in theattachment area was usually 4-10 particles thick. Macro-phage pseudopods were seen in contact particularly withCF-labelled parasite regions (Fig. 8). CF removal bymacrophages was observed after the attachment phase(Fig. 10). On the contrary, the attachment betweenmacrophages and amastigotes or infective promastigoteswas mediated by unlabelled focal membrane areas(Figs 14 and 15). Later these parasite forms were seen inCF-free endocytic vacuoles.

After interaction, many macrophages presentedvacuoles containing CF-labelled vesicles. Small vacuoles(Fig. 11) probably gave rise to larger ones (Fig. 12) byvacuole fusion. Extracellular CF-labelled membraneswere often seen attached to macrophages (Fig. 11).

Electrophoretic mobility of parasites before and after CFtreatmentAll the evolutive forms of L. mexicana amazonensisstudied exhibit negatively charged surfaces, since theparasites migrate, in random orientation, towards thenegative electrode under an applied voltage.

The average mobilities of the parasites are closelysimilar. Amastigotes have a mean EPM value of— l-12|iwns~ V~ cm, while infective and non-infectivepromastigotes have mean EPM values of —1-12 and— l-14,ums~ V" 'cm, respectively. As can be deducedfrom the populational analysis shown in Fig. 16, the threeparasite forms studied by cell electrophoresis consist ofpooled cell suspensions with different EPM values. Themodal EPM values fall into the class interval of —1-0 to— l-2[ims~l V"1 cm for amastigotes, and non-infective

484 E. M. B. Saraiva et al.

P

/T7

v !

•' V,

8

10

and infective promastigotes. Incubation of the parasitesin the presence of cationized ferritin render their surfacesless negative. The mean EPM values decreased slightly(12%) after treatment of amastigotes and infective pro-mastigotes with the cationic particles. On the other hand,

? l"J12

in non-infective promastigotes the mean EPM decreasedby about 26% after CF treatment. However, as can beseen in Fig. 16, this treatment was effective in alteringthe modal EPM values of the parasites. The morefrequent EPM values are now found between — 0"8 and

Leishmania-w«r/o/>/;«£fe interaction 485

1

14

13

<* 1.-

• * %

15

Fig. 13. Macrophage showing several endocytosed non-infective promastigotes. Large amounts of bound CF can be seen oninteriorized parasites. X 17 000.Fig. 14. Discontinuous attachment area between infective promastigote and macrophage. Parasite cells are completely devoid ofCF. X30 000.Fig. 15. Focal attachment between amastigote and macrophage. X52 000.

Amastigotes Non-infective Infective100,

50

0/100

50

0

100EPM = -l-12/ims"'v"'cm

CF-treated

50

0/100

EPM = -0-99/«ns~1v-1cm

50

0

promastigotes JQQ. promastigotes

EPM= - l 'v 'cm EPM = — 1-12 funs v cm

_1—1 50

CF-treated•0/100

EPM = -0-84 fan s~' v"1 cm

50

CF-treated

EPM= -0-99^ms"'v"1cm

-0-7 -1-4

486 E. M. B. Saraiva et al.

-0-7 -1-4 - 0 - 7 -1-4

Fig. 16. Populational analysis of theelectrophoretic mobilities of differentdevelopmental L. m. amazonensis formsbefore (top) and after (bottom) treatmentwith 10 fig ml"1 cationized ferritin (CF).EPM = mean value of electrophoreticmobility.

— 1-0/ims V cm. Indeed, new mean class intervals ofnon-infective and infective promastigotes could be ob-tained after measurement of the EPM of parasites fromCF-treated populations. These mean class intervals werefound between —0-6 and —0-8/ims"1 V"1 cm for bothnon-infective and infective promastigotes.

Discussion

Our present observations show CF-induced capping andshedding of anionic sites from L. mexicana amazonensis.CF rearrangement has been described in several celltypes including a free-living protozoan (King & Preston,1977), mycoplasma (Schiefer et al. 1976) and sporozoanparasites (Augustine & Danforth, 1984; Russell & Sin-den, 1981), but to date there are no reports on CF-induced movement of surface ionogenic sites on trypano-somatids. This may be due to the fact that fixed cells werealways used for CF labelling. In previous work nodifference in CF-labeling patterns between living andglutaraldehyde-fixed promastigotes of Leishmania dono-vani, L. tropica and L. braziliensis (Muhlpfordt, 1975)and Trypanosoma brucei, T. gambiense and T. rho-desiense (Muhlpfordt, 1977) was observed. Such resultsmay be explained by the probable use of agglutinating CFconcentrations, as agglutination was shown to inhibitantibody-induced capping in L. donovani (Dwyer, 1976)and T. cruzi (Leon et al. 1979; Schmunis et al. 1980).

The CF-induced capping phenomenon is similar inmany aspects to other described receptor-ligand systemsin Leishmania, such as specific antibodies (Doyle et al.1974; Dwyer, 1976) and lectins (Doyle et al. 1974). Itdisplays remarkable movement polarity of the CF-aggre-gated sites towards the flagellated end of the parasites.Antibody- or lectin-induced cap formation in L. ennettii(Doyle et al. 1974), L. donovani (Dwyer, 1976) and T.cruzi (Szarfman et al. 1980; Schmunis et al. 1980) wasobserved at both poles of the cells, but generally at thatpole where the flagellar pocket is found, i.e. the anteriorin Leishmania, and the posterior in T. cruzi trypomasti-gotes (Dwyer, 1976; Szarfman et al. 1980; Schmunis etal. 1980). In our preparations, CF particles were alsousually observed aggregated at the anterior (flagellated)portion of the cells and CF labelling was rarely seen onposterior ends. Antibody-induced shedding in L. dono-vani (Dwyer, 1976) and T. cruzi (Schmunis et al. 1980)usually takes place in the basal region of the flagellum,close to the flagellar pocket. It is noteworthy that theflagellar pocket of trypanosomatids is generally involvedin several membrane-exchanging phenomena includingendocytosis and exocytosis (reviewed by De Souza, 1981;Coppensefa/. 1987).

Cationized ferritin redistribution in unfixed baby ham-ster kidney cells (Grinnel et al. 1975) and Ehrlich ascitestumor cells (Subjeck & Weiss, 1975) did not lead to theformation of defined caps, and in the first case it was notblocked by metabolic inhibitors, suggesting that theinitial phase of this process is brought about by asymmet-ric lateral diffusion of cross-linked anionic sites. In L.mexicana amazonensis the formation of defined CF caps

was energy-dependent and followed by liberation oflabelled membranes, similar to the antibody-inducedshedding in T. cruzi (Schmunis et al. 1980) and theConA-ferritin shedding in L. braziliensis (Ayesta et al.1985). Skutelsky & Danon (1976) suggested that surfaceanionic sites were detached from endothelial cells withclustered CF particles, although associated membranescould not be seen. In L. mexicana amazonensis shedmembrane units were hardly seen without altering thepreparation angle, except when low azide concentrationswere used at 4°C to slow down the process, presumablyby diminishing the amount of CF bound to detachingmembranes, which were then less twisted.

In contrast to lectin and antibody-induced capping,CF-induced capping in L. mexicana amazonensis wasnot inhibited at 4°C. Low temperatures also failed toinhibit antibody-induced capping of host immunoglobu-lin G (IgG) on T. lewisi (Giannini & D'Alesandro, 1978),CF-induced reorganization in Naegleria (King & Pres-ton, 1977) and rat thyroid cells (Nishiyama et al. 1982).

Although CF-binding could be related to infectivity inL. braziliensis (Ayesta et al. 1985), we observed nodifference between CF-binding to infective and non-infective cells. However, it must be noted that infectivecells shed CF-binding sites more quickly than non-infective ones, being almost completely devoid of CFafter 30min. These data suggest that L. mexicanaamazonensis promastigotes may have different surfaceproperties concerning the mobility of anionic sites. Therole of membrane fluidity and surface component mo-bility on parasite infectivity remains to be clarified but itis interesting to note that opsonized T. cruzi trypomasti-gotes of the Y strain, which are able to survive inmacrophages, can readily cap bound antibodies, whereasthe CL strain, which takes a longer time to induce capformation, is destroyed inside the phagocytes (Schmuniset al. 1980; Alcantara & Brener, 1978).

CF treatment of non-infective promastigotes produceda remarkable increase in their uptake by macrophages,and a reduction in their electrophoretic mobility, prob-ably because of its slower shedding. Therefore, theremaining bound CF could reduce the negative surfacecharge of these cells, reducing the electrostatic repulsionbetween the parasite and the macrophage surfaces, asattachment was mediated by CF-labelled membranes.When two macrophages were attached to a parasite, CFcould be seen solely in attachment areas (Fig. 7). SeveralLeishmania species (reviewed by Chang, 1983) and L.mexicana amazonensis promastigotes, particularly(Pimenta & De Souza, 1986), were shown to interact withmacrophages predominantly at the flagellar end first.Cationized ferritin-treated non-infective promastigotesalso presented CF in the flagellar attachment (Fig. 9),suggesting that anionic sites can also take part in theinitial contact between parasite and phagocyte. As amas-tigotes and infective promastigotes quickly removedbound CF, its effects on EPM and uptake by macro-phages were not significant. Meirelles et al. (1984)showed that CF treatment of T. cruzi cells reduced theEPM and increased the uptake of trypomastigoteswhereas epimastigotes had neither EPM nor endocytic

heishmama-macmphage interaction 487

indexes altered by CF. Once again, charge reduction isshown to promote increased endocytosis.

Pimenta & De Souza (1988) demonstrated that CF andparasites were found in different vacuoles after macro-phage-L. mexicana amazonensis interaction in the pres-ence of CF, suggesting different endocytic pathways.Our results showing CF removal from prelabeled non-infective promastigotes also indicate that different mech-anisms are used for their uptake by the macrophages.

Evidence has been accumulated on the role of naturalor synthetic polycationic agents during the interaction ofpathogens with their hosts, such as in the cases of thepenetration of Toxoplasma gondii into host cells (Werk etal. 1984) and the bacterial phagocytosis by polymorpho-nuclear leukocytes and macrophages (Pruzanski & Saito,1978).

Two different mechanisms could explain the increasedphagocytosis of CF-treated cells. First, CF moleculesmay function as cross-linkers between negatively chargedmembranes of the interacting cells. Second, bound CFmay reduce the negative surface charge and thus permitclose apposition of negatively charged cells. Previousreports (Mutsaers & Papadimitriou, 1988; Denef &Ekholm, 1980) suggest that the second mechanism ismore likely when the CF layer is more than one particlethick. As demonstrated (Capo et al. 1981), neuramini-dase treatment and poly-L-lysine can increase erythrocytebinding to macrophages and produce the formation oflarge continuous attachment areas. Therefore, chargereduction or neutralization may have similar effects onparasite attachment to macrophages. Larger attachmentareas presenting a layer four to ten particles thick, as wellas an enhanced endocytic rate, strongly suggest the role ofsurface charge reduction in the uptake of CF-treated non-infective promastigotes. Mouse peritoneal macrophageswere shown (Skutelsky & Hardy, 1976) to remove boundCF by endocytosis and surface detachment and to take afew hours to recover the CF binding. Therefore, macro-phage CF treatment, producing removal of surface an-ionic sites, could increase non-infective promastigoteuptake by reducing electrostatic repulsion. It remains tobe explained why the other parasite forms were notaffected by the same procedure.

The electrokinetic behavior of the parasite populationsrevealed interesting aspects. Both infective and non-infective promastigotes represent highly heterogeneouspopulations, since after treatment with cationized ferritinnew subpopulations of both promastigotes could beobserved. Thus, it is possible that the results of interac-tion experiments could be due much more to intrinsicvariations in a population than merely to the bindingefficiency of a non-infective population when comparedwith the infective one. The same reasoning must beapplied to amastigotes.

Detailed knowledge about the chemical nature of thesurface anionic groups of L. mexiccma amazonensis isrequired for a better understanding of the interactionprocess. This subject is under investigation.

The authors thank Dr H. Meyer and Dr Paulo F. P. Pimentafor reading the manuscript, Mr Antonio de Oliveira for

technical assistance and Mrs Alba Valeria Peres for secretarialassistance. This work has been supported by FINEP, CNPqandCEPG-UFRJ.

References

ALCANTARA, A. & BRENER, Z. (1978). The in vitro interaction ofTrypaiiosoma cnizi bloodstream forms and mouse peritonealmacrophages. Ada tropica 35, 209-219.

AUGUSTINE, P. C. & DANFORTH, H. D. (1984). Effects of cationizedferritin and neuraminidase on invasion of cultured cells by Eimeriameleagrimitis sporozoites. J. Pmtozool. 31, 140-144.

AYESTA, C , ARGUELLO, C. & HERNANDEZ, A. G. (1985).

Leishmania brazilieiisis: cell surface differences in promastigotes ofpathogenic and nonpathogenic strains. Expl Parasitol. 59, 185-191.

CAPO, C , BONGRAND, P., BENOLIEL, A. M., RYTER, A. &: DEPIEDS,

R. (1981). Particle-macrophage interaction: role of surfacecharges. Amils Immunol. (Inst. Pasteur) 132D, 165-173.

CHANG, K. P. (1983). Cellular and molecular mechanisms ofintracellular symbiosis in Leishmaniasis. Int. Rev. Cvtol. Suppl. 14,267-305.

COPPENS, I., OPPERDOES, F. R., COURTOY, P. J. & BAUDHUIN, P.

(1987). Receptor-mediated endocytosis in the bloodstream form ofTrypanosoma brucei.J. Pmtozool. 34, 465-473.

DANON, D., GOLDSTEIN, L., MARIKOVSKY, Y. & SKUTELSKY, E.

(1972). Use of cationized ferritin as a label of negative charges ofcell surfaces. J . Ultrastntct. Res. 38, 500-510.

DENEF, J. F. & EKHOLM, R. (1980). Membrane labeling withcationized ferritin in isolated thyroid follicles. J. Ultrastnict. Res.71, 203-221.

D E SOUZA, W. (1981). Cell Biology of Trypanosoma cmzi. Int. Rev.Cytol. 86, 197-283.

D E SOUZA, W., ARGOELLO, C , MARTINEZ-PALOMO, A., TRISSL, D.,

GONZAL£Z-ROBLES, A. & CHIARI, E. (1977). Surface charge ofTiypanosoma cnizi. Binding of cationized ferritin andmeasurement of cellular electrophoretic mobility. J'. Pmtozool. 24,411-415.

DOYLE, J. J., BEHIN, R., MAUEL, J. & ROWE, D. S. (1974).

Antibody-induced movement of membrane components ofLeishmania enriettii.J. exp. Mecl. 139, 1061-1069.

DWYER, D. M. (1976). Antibody-induced modulation of Leishmaniadonovani surface membrane antigens. J. Iminun. 117, 2081-1091.

GIANNINI, M. S. & D'ALESANDRO, P. A. (1978). Unusual antibody-induced modulation of surface antigens in the cell coat of abloodstream trypanosome. Science 201, 916-918.

GRINNEL, F., TOBLEMAN, M. Q. & HACKENBROCK, C. R. (1975).The distribution and mobility of anionic sites on the surfaces ofbaby hamster cells. J . Cell Biol. 66, 470-479.

KING, C. A. & PRESTON, T. M. (1977). Studies of anionic sites onthe cell surface of the amoeba Naegleria gmbeii using cationizedferritin. J. Cell Sci. 28, 133-149.

LEON, W., VILLALTA, F., QUF.IROZ, T. & SZARFMAN, A. (1979).Antibody-induced capping of the intracellular stage ofTiypanosoma cnizi. Infect. Immun. 26, 1218-1220.

MATSUI, H., ITO, T. & OHNISHI, S. (1983). Phagocytosis bymacrophages. III. Effects of heat-labile opsonin and poly-L-lysine.J. Cell Sci. 59, 133-143.

MEIRELLES, M. N. L., SOUTO-PADR6N, T. & DE SOUZA, VV. (1984).

Participation of cell surface anionic sites in the interaction betweenTiypanosoma cnizi and macrophages. J. sitbmicmsc. Cvtol. 16,533-545.

MOSSER, D. M., VLASSARA, H., EDELSON, P. J. & CERAMI, A.

(1987). Leishmania promastigotes are recognized by themacrophage receptor for advanced glycosylation endproducts.J. exp. Mecl. 165, 140-145.

MilHLPFORDT, H. (1975). Vergleichende elektronenmikroskopischeUntersuchung fiber die Markierung von I^eishmania donovani, L.tropica und L. brazilieiisis mit Ferritin. Tropenmed. Parasit. 26,385-389.

MUHLPFORDT, H. (1977). Cell surface labelling of bloodstreamtrypanosomes with cationic ferritin. Protozoology 3, 71-74.

MUTSAERS, S. E. & PAPADIMITRIOU, J. M. (1988). Surface charge of

488 E. M. B. Saraiva et al.

macrophages and their interaction with charged particles. J.leukocyte Biol. 44, 17-26.

NISHIYAMA, F., TAKATA, K., FUKUDA, M. & HIRANO, H. (1982).

Regional differences in surface anionic sites in the thyroid folliclecells shown by cationized ferritin labelling. Okajimas Folia anal.jfpn. 58, 289-304.

PAPADIMITRIOU, J. M. (1982). An assessment of the surface charge ofsingle resident and exudate macrophages and multinucleate giantcells. J . Path. 138, 17-24.

PIMENTA, P. F. P. & DE SOUZA, \V. (1983). Leishamania mexicanaamazonensis: surface charge of amastigote and promastigote forms.Expl Parasitol. 56, 194-206.

PIMENTA, P. F. P. & DE SOUZA, W. (1986). Structural observationson the attachment of promastigotes of L. mexicana amazonensis tothe surface of macrophages. Microsc. Elelron. Biol. Cell 10, 65-75.

PIMENTA, P. F. P. & DE-SOUZA, W. (1988). Freeze-fracture andcytochemistry study of the interaction between Leishmaniamexicana amazonensis and macrophages. J. snbmicrosc. Cvtol.Path. 20, 89-99.

PRUZANSKI, W. & SAITO, S. (1978). The influence of natural andsynthetic cationic substances on phagocytic activity of humanpolymorphonuclear cells. An alternative pathway of phagocyticenhancement. Expl Cell Res. 117, 1-13.

PUENTES, S. M., SACKS, D. L., DA SILVA, R. P. & JOINER, K. A.

(1988). Complement binding by two developmental stages ofleishmania major promastigotes varying in expression of a surfacelipophosphoglycan. J. exp. Med. 167, 887-902.

RUSSEL, D. G. & WRIGHT, S. D. (1988). Complement receptor type3(CR3) binds to an Arg-Gly-Asp-containing region of the majorsurface glycoprotein, gp63, of Leishmmania promastigotes. J. exp.Med. 168," 279-292.

RUSSELL, D. G. & SINDEN, R. E. (1981). The role of thecytoskeleton in the mobility of coccidian sporozoites. J. Cell Sci.50, 345-359.

RUSSELL, D. G. & WILHELM, I-I. (1986). The involvement of GP63,the major surface glycoprotein in the attachment of Leishmaniapromastigotes to macrophages. J'. Immnn. 136, 2613-2620.

SANTOS, A. B. S. & DE SOUZA, W. (1983). Surface charge andultrastructure of the cell surface of resident and thioglycolate-elicited mouse peritoneal macrophages. J. snbmicrosc. Cvtol. 15,897-911.

SARAIVA, E. M. B., ANDRADE, A. F. B. & DE SOUZA, W. (1987).Involvement of the macrophage mannose-6-phosphate receptor inthe recognition of Leishmania mexicana amazonensis. Parasil.Res. 73, 411-416.

SARAIVA, E. M. B., ANDRADE, A. F. B. & PEREIRA, M. E. A. (1986).A comparative study of infective and non-infective forms ofleishmania mexicana amazonensis. Eur.J. Cell Biol. 40, 219—225.

SARAIVA, E. M. B., PIMENTA, P. F. P., PEREIRA, M. E. A. & DE

SOUZA, W. (1983). Isolation and purification of amastigotes ofleishmania mexicana amazonensis by a gradient of metrizamide.J. Parasil. 69, 627-629.

SCHIEFER, H. G., KRAUSS, H., BRUNNER, H. & GERHARDT, U.

(1976). Ultrastructural visualization of anionic sites onMycoplasma membranes by polycationic ferritin. J. Bad. 127,461-468.

SCHMUNIS, G. A., SZARFMAN, A., D E SOUZA, \V. & LANGEMBACH,

T. (1980). Tiypanosoma cmzi: Antibody-induced mobility ofsurface antigens. Expl Parasit. 50, 90-102.

SILVA-FILHO, F. C , SANTOS, A. B. S., CARVALHO, T. M. U. & D ESOUZA, W. (1987). Surface charge of resident, elicited andactivated mouse peritoneal macrophages. J'. Leukocyte Biol. 41,143-149.

SKUTELSKY, E. & DANON, D. (1976). Redistribution of surfaceanionic sites on the luminal front of blood vessel endothelium afterinteraction with polycationic ligand.J'. Cell Biol. 71, 232-241.

SKUTELSKY, E. & HARDY, D. (1976). Regeneration of plasmalemmaand surface properties in macrophages. Expl Cell Res. 101,337-345.

SOUTO-PADR6N, T. & DE SOUZA, W. (1986). The surface charge ofTiypanosoma crnzi: analysis using cell electrophoresis, lectins andultrastructural cytochemistry. J. snbmicrosc. Cvtol. 18, 701-709.

SUBJECK, J. R. & WEISS, L. (1975). The binding of cationizedferritin at the surfaces of Ehrlich ascites tumor cells: the effect ofpH and glutaraklehyde fixation. J . cell. Physiol. 85, 529-536.

SZARFMAN, A., QUEIROZ, T. & D E SOUZA, W. (1980). Mobility of

Concanavahn A receptors in Tiypanosoma cmzi. J. Parasitol. 66,1055-1057.

TuRCO, S. J. (1988). The lipophosphoglycan of Leishmania. Parasil.Today 4, 255-257.

VAN OSS, C. J. (1978). Phagocytosis as a surface phenomenon.A. Rev. Microbiol. 32, 19-39."

WERK, R., DUNKER, R. & FISCHER, S. (1984). Polycationic

polypeptides: a possible model for the penetration-enhancing factorin the invasion of host cells by Toxoplasma gondii. J. i>en.Microbiol. 130, 927-933.

WILSON, M. E. & PEARSON, R. D. (1986). Evidence that leishmaniadonovani utilizes a mannose receptor on human mononuclearphagocytes to establish intracellular parasitism. J'. Immnn. 12,4681-4688.

WOZENCROFT, A., SAYERS, G. & BLACKWELL, J. M. (1986).Macrophage type 3 complement receptors mediate serum-independent binding of Leishmania donovani. J. exp. Med. 164,1332-1340.

WYLER, D. J., SPYEK, J. P. & MCDONALD, J. A. (1985). In vitroparasite-monocyte interactions in human leishmaniasis: possiblerole of fibronectin in parasite attachment. Infect. Immnn. 49,305-309.

(Received 26 January I9S9 - Accepted 6 April 1989)

Leishmama-macrophage interaction 489