Red cell membrane protein distribution during malarial invasion

A. R. DLUZEWSKI1, P. R. FRYER2, S. GRIFFITHS2, R. J. M. WILSON3 and W. B. GRATZER1

^Medical Research Council Cell Biophysics Unit, King's College, 26-29 Dairy Lane, London WC2B 5RL, UK2Clinical Research Centre, Watford Road, Harrow, Middlesex HAl 3UJ, UK^National Institute for Medical Research, The Ridgeway, Mill Hill, London N\V7 1AA, UK

Summary

Immuno-gold labelling electron microscopy of thinsections was used to determine the distribution ofred cell membrane and membrane skeleton pro-teins in the vicinity of internalized malaria para-sites. When examined immediately after invasion(young ring-stage parasites), the parasitophorousvacuole membranes of both Plasmodium falci-parum and P. knowlesi were found to be character-ized by the essentially complete absence of spec-trin, ankyrin and the most abundant trans-membrane protein, band 3. P. knowlesi merozoiteswere trapped in the attached but not internalizedstate by pretreatment with cytochalasin B. In thismerozoite-red cell complex antibody labellingshowed that band 3 had been eliminated from theregion of the host cell membrane in contact with theparasite. Internal vesicles, originating apparently

from the site of attachment, were often observed inthe red cell. Opposite the attached parasite a cavitywas also sometimes seen in the host cell, presum-ably representing an Incipient internal vesicle. Themembrane was intact, as judged by the absence ofprotein (haemoglobin) in the cavity, and, like themembranes surrounding the internal vesicles, wasdevoid of membrane proteins. A large multilamel-lar body was sometimes seen in the merozoite closeto its point of attachment. The lamellar spacing wasabout 50 nm. The electron microscope imagessuggest a diffusion of electron-dense material fromthe lamellar body into the cavity in the host cell.

Key words: malaria, erythrocyte membrane, immunoelectronmicroscopy.

Introduction

The sequence of events involved in the penetration of themalaria parasite into the red cell have been defined by thestudies of Aikawa, Miller and their colleagues (Aikawa etal. 1978, 1981). Following recognition and attachment tosurface receptors, invagination of the host cell membraneoccurs and the parasite moves into the cell. At an earlystage in this process it appears that intramembraneparticles of the host cell collect in an electron-dense ringat the junction with the parasite. By contrast, theinvaginated membrane, opposite the apex of the parasite,becomes depleted of particles (McLaren et al. 1979;Aikawa et al. 1981). The simplest model for invasion isthat the parasite encapsulates itself in the host cellmembrane bilayer, but it has also been suggested (Ban-nister et al. 1986) that the parasitophorous vacuolemembrane may be synthesized by the parasite.

The composition of the parasitophorous vacuole mem-brane is unknown. Recently, Atkinson et al. (1987) haveshown that the membranes of mature intraerythrocyticparasites (mature rings, trophozoites and schizonts) aredevoid of spectrin, band 3 and glycophorin A. In thesestages of development, however, extensive structural

Journal of Cell Science 92, 691-699 (1989)Printed in Great Britain © The Company of Biologists Limited 1989

changes and, very possibly, proteolysis have occurred.We have accordingly examined freshly invaded cells,containing either Plasmodium falciparum or more par-ticularly Plasmodium knowlesi, in which synchronousinvasion can be much more readily achieved. We have, inaddition, studied cells to which P. knowlesi merozoitesare attached only at the outer surface. Our results indicatethat the red cell membrane proteins, both integral andskeletal, are eliminated from the area of attachment of theparasite and are absent from the parasitophorous vacuoleof the freshly internalized parasites. The observations donot exclude an erythrocytic origin for the parasitophorousvacuole membrane.

Materials and methods

Parasite culturesP. falciparum parasites were cultured in vitro by the method ofTrager & Jensen (1976), and synchronized by the sorbitolprocedure (Lambros & Vandenberg, 1979). P. knowlesi para-sites were obtained from a Rhesus monkey, that had beeninoculated with material from a frozen stabilate. Invasion wasinduced by the addition of purified schizont preparations(Dluzewski et al. 1984) to human or simian red cells in an

691

RPMI 1640 culture medium, containing 10% human serum forexperiments with P. falciparum and foetal calf serum for P.knowlesi. Externally attached P. knotvlesi merozoites resultedfrom the following procedure: purified mature schizonts wereagitated in suspension at room temperature with a vortex mixerat intervals of IS s for 1 min. Cytochalasin B (Sigma) was addedfrom a 10 nig ml" stock solution in dimethylsulphoxide to thefree merozoites and rupturing schizonts to give a concentrationof 2/igml"1. After 3 min at room temperature the target cellswere added and the suspension was kept at 37 °C in anatmosphere of 5% oxygen, 7% carbon dioxide and 88%nitrogen.

AntibodiesPolyclonal antibodies against human spectrin and ankyrin weregenerated in rabbits, using the purified proteins prepared bystandard procedures (Gratzer, 1982; Bennett & Stenbuck,1980, respectively) as antigens. They were tested for specificityby Ouchterlony diffusion and by staining electroblots ofSDS-polyacrylamide gels. Sheep antibodies against rabbitband 3 were produced and characterized as described byFoxwell & Tanner (1981). They are directed against thecytoplasmic domain of the protein and react efficiently withhuman band 3. Control antisera were obtained by absorbingantisera overnight at 4°C with purified red cell membranes.These were prepared by freezing and thawing the cells,suspending them in 15 vols of 5 mM-potassium phosphate,pH 7-4, and centrifuging the suspension at 42 000g. The ghostswere twice washed in the same buffer at 4°C.

Electron microscopy'For embedding with Lowicryl K4M resin, washed cells werelysed by addition of 0-1 mg ml" saponin in phosphate-bufferedisotonic saline (PBS). The resulting ghosts were recovered bycentrifugation at 11 000 g for 3 min and washed three times withPBS. They were then fixed with 1 % glutaraldehyde in PBS,pH7-4, for 30 min at room temperature. After three furtherwashes with PBS, the ghosts were dehydrated by successiveimmersion in five solutions of increasing methanol concen-tration up to 90% (v/v). The dehydrated ghosts were embed-ded in the Lowicryl K4M, following Roth et al. (1981). Thetemperature was reduced from ambient to —20°C beforepolymerization of the resin. Thin sections were cut (Reichertultramicrotome) and mounted on Formvar-coated 200-meshcopper grids.

For embedding with LR White resin, intact cells werewashed with 0-1 M-potassium phosphate, p H 7 4 , and fixed bysuspension in 0-25 % glutaraldehyde in the same buffer for20 min at room temperature (Brown e( a/. 1985). After additionof 50mM-ammonium chloride, 0-1 M-potassium phosphate, toeliminate the remaining glutaraldehyde, followed after 20 minby one further wash in this medium, the cells were dehydratedby successive immersion in five solutions of increasing ethanolcontent up to 75%. The samples were equilibrated with LRWhite monomer, transferred to gelatin capsules and the resinwas allowed to polymerize at 50°C for 24 h. Sections were cutand mounted as before, but without Formvar.

ImmunofluorescenceRed cells attached to slides were lysed with saponin in isotonicsaline, or air-dried smears were permeabilized with ice-coldacetone for 15 min or methanol for 10 s at room temperature.The latter were rinsed with distilled water and incubated withanti-spectrin or anti-band 3, for 1 h at 37 °C in a moist chamber.They were rinsed three times with PBS and incubated at 37°Cfor 1 h followed with Rhodamine-labelled goat anti-rabbit or

fluorescein-labelled rabbit anti-sheep IgG (Miles). After threerinses with PBS and one with distilled water, the slides were air-dried and examined in a Zeiss microscope, equipped forepifluorescence.

Immuno-electron microscopyGrids were floated on 50mM-potassium phosphate, pH7-4,containing 2-5 rngml" Tween 20 and 20mgml~ bovine serumalbumin, for 5 min at room temperature. They were floated for1 h on a meniscus of antibody solution diluted between 1:10 and1: 30 with the same medium, washed three times by flotation onthe antibody-free buffer, and incubated at room temperaturewith a 1:10 (v/v) dilution of a preparation of colloidal gold-labelled protein A, produced by conjugating protein A (Sigma)with chlorauric acid (BDH), as described by Roth (1982). Thegrids were washed twice with the buffer, stream-washed withdistilled water, air-dried, and stained with uranyl acetate andlead citrate. LR White sections were coated with carbon.Sections were examined in a Philips 300 electron microscope at80 kV accelerating voltage.

Results

We attempted to detect host cell membrane proteins inthe parasitophorous vacuole by immunofluorescence,using rabbit anti-spectrin and sheep anti-rabbit-band 3antibodies, in association with the appropriate fluor-escent second antibody. The plasma membranes ofinfected and uninfected red cells gave identical brightfluorescence. No additional intensity could be discernedfrom the region of the parasite in cells that had been lysedwith saponin or permeabilized with acetone or withmethanol. This suggested that the parasite could havecontained little spectrin or band 3, but because of thedominant, signal from the host cell membrane (even whensaponin lysis had been used to eliminate haemoglobin),the method is clearly too insensitive to permit any but thecrudest qualitative conclusions. We accordingly resortedto immuno-electron microscopy to obtain quantitativeinformation.

Fig. 1A shows the results of labelling sections ofhuman red cells, containing the young ring form of P.falciparum, with anti-spectrin antibodies. Labelling ismuch more abundant in the lysed cells, embedded inLowicryl (Fig. 1A, inset), than in sections of intact cellsin LR White. This may be the result of denaturation ofprotein epitopes in the course of thermal polymerizationof the LR White (Brown et al. 1985), and/or an occlusiveeffect of the haemoglobin in contact with the antigenicsite. In the saponin-lysed cells the characteristic cellcontour is lost, but the membrane is clearly identified.The parasitophorous vacuole membrane is not disruptedby saponin treatment (Siddiqui et al. 1979). In bothsystems labelling by anti-spectrin is seen on the red cellmembrane with greater or lesser abundance, but noantigen is ever detected on the parasitophorous vacuolemembrane. A similar distribution of antigen was ob-served with anti-ankyrin antibodies (Fig. IB). We thusconclude that the membrane cytoskeleton of the host cellis excluded from the parasitophorous vacuole. Similarstudies were carried out with an antibody against thecytoplasmic domain of the preponderant transmembrane

692 A. R. Dluzewski et al.

1A

:IV r

B

Fig. 1. Immuno-gold-labelled sections of human red cells, parasitized by P. falciparum. A. Intact cells, embedded in LR Whiteresin and labelled with a polyclonal anti-spectrin antibody, followed by gold-labelled protein A. Note labelling on the red bloodcell membrane but none at the periphery of the ring-form parasite (r). Inset: parasitized cells, lysed with saponin, embedded inLowicryl K4M and similarly labelled. B. Sections of parasitized cells equivalent to those shown in A, but labelled with apolyclonal anti-ankyrin antibody, showing again absence of gold label on the ring-stage parasites (r). Bars, 1 ^m.

Membrane proteins in parasitized red cell 693

\

2A

- \

B

w

\

Fig. 2. Sections of P. falcipa rum-parasitized red cells, labelled with anti-band 3 antibodies followed by gold-labelled protein A.A. This shows a rare example of an unlysed cell that has survived saponin treatment prior to embedding in Lowicryl K4M.B. Saponin-lysed cell, similarly embedded and labelled; note labelling only on the red cell membranes. C. Control, exposed tomembrane-absorbed antibody, followed by gold-labelled protein A. r, ring-stage parasites. Bars, 1 Jim.

694 A. R. Dluzezvski et al.

Fig. 3. Sections of P. ^now)/w;-parasitized monkey red cells, shortly after invasion: A, labelled with anti-spectrin antibodies,followed by gold-labelled protein A; and B, labelled with anti-band 3 antibodies, followed by gold-labelled protein A. r, ring-stage parasite; m, merozoite. Bars, 1 )im.

Membrane proteins in parasitized red cell 695

Table 1. Distribution of anti-band 3 antibodies on thin sections of P. knowlesi-infected red cells

Material

Parasitized cells

Cells exposed to cytochalasinB-treated parasites

•Standard deviation.

Membrane examined

Host cell membraneParasitophorous vacuole

membrane

Host cell membraneHost cell membrane in

contact with parasiteInternal vacuoles

Gold particles per /tmmembrane section

2-31 ±0-80*0-08 ±0-22

3-47 ± 1-070-37 ±0-96

0-00 ±0-00

Number of cellsanalysed

1111

2515

10

Total lengthof membraneanalysed (/un)

18389

16920

23

protein, band 3. Abundant gold labelling was observedon the host cell membrane, but again no band 3 could beseen on the parasitophorous vacuole (Fig. 2). Cells werelysed in this system before sample preparation to pre-clude inhibition of photo-induced polymerization byhaemoglobin. Fig. 2A shows a rare instance of a parasit-ized, unlysed cell. As with the other antisera, the controlserum showed no labelling of the host cell or parasitemembrane (Fig. 2C).

The experiments with anti-spectrin and anti-band 3were repeated on sections of newly invaded monkey redcells, containing P. knowlesi. These cells were fixed40 min after exposure of the target cells to a highlysynchronous inoculum of late-stage schizonts. Again(Fig. 3) no significant degree of labelling could be seen onthe parasitophorous vacuole, indicating that the mero-zoite carries little or no host cell membrane proteins intothe cell with it. In this system the age of the rings isbetween 0 and 40min, compared with, say, 4h for P.falciparum. The likelihood of disappearance of antigensthrough proteolytic destruction is therefore muchreduced. A quantitative analysis of the distribution of thelabelled band 3 was undertaken. Sampling separateelectron micrographs, the numbers of gold particles perunit length of membrane in the section were counted.The results are given in Table 1, from which it followsthat the content of band 3 in the parasitophorous vacuolemembrane is less that 4% of that of the red cellmembrane.

We have attempted to determine the abundance of thehost cell protein in the region of attachment of theparasite before its passage into the cell. Parasites weretreated with cytochalasin B to arrest the interaction withthe red cell at the stage of attachment, following aprocedure modified from that of AikawaeJ al. (1981). Wefound that such treatment did not impede invasion by P.falciparum, but with P. knowlesi considerable numbersof attached merozoites could be observed after fixation(Fig. 4). Labelling with anti-band 3 antibody revealed anextensive diminution in band 3 concentration in the zoneof contact between the plasma membrane and the at-tached merozoite (Fig. 4A and B, and Table 1). Thusattachment causes or is accompanied by local disappear-ance of. the transmembrane protein. With anti-spectrinwe also observed no labelling in the area of contact, butbecause of the sparser distribution of label generally, the

particle counts were insufficient for a statistical analysis.The attached, cytochalasin-treated merozoites also

caused the formation of haemoglobin-free vacuoles insidethe host cell, originating apparently from the site ofcontact. These are presumably identical to the 'secondaryvacuoles' of Aikawa et al. (1981). The membrane sur-rounding these structures was devoid of band 3 (Fig. 4A(inset), and Table 1). Some of the attached parasiteswere seen to contain laminar structures, made up ofconcentric electron-dense layers with a spacing of roughly50 nm. Fig. 4C shows an instance of the apparent devel-opment of an internal vacuole from the site of attach-ment; material from or near the laminar body appears tobe diffusing into the cavity in the host cell.

Discussion

Atkinson et al. (1987) have shown that the late erythrocy-tic stages of P. falcipaivm are marked by an absence ofspectrin, glycophorin A and band 3 from the parasito-phorous vacuole membrane, but they observed reactivityagainst anti-ankyrin in Maurer's clefts in the parasitizedcell. We showed in a preliminary note (Dluzewski et al.1988) that spectrin was absent from the parasitophorousvacuole of young ring-stage P. falciparum. We have nowdemonstrated that in P. falciparum, and especially in P.knowlesi, in which the parasites are only between 0 and40min old, the host cell proteins are likewise excludedfrom the parasite, observed directly after invasion. Ourresults are thus independent of the changes, proteolyticand other, that accompany the maturation of the internal-ized parasite beyond the young ring-stage. Moreover,unless the attachment of the cytochalasin-treated para-sites is a radically different process from that leading tothe 'irreversible' state of attachment in the normal courseof invasion, then the process of elimination of the hostcell proteins from the membrane at the site of entrybegins before the parasite starts its passage into the cell.The coupling between the movement of transmembraneproteins and the membrane cytoskeleton has been notedfrequently (Nicolson & Painter, 1973; Ji & Nicolson,1974); for example, at sites of incipient endocytosis, bothintramembrane particles (predominantly band 3) and thecytoskeletal network are withdrawn (Hardy et al. 1979).Such processes have also been reported to be linked tochanges in spectrin phosphorylation (Gazith et al. 1976;

696 A. R. Dluzewski et al.

m

m

4A

.;#/T7

Fig. 4. Sections of cytochalasin B-treated P. knotolesi merozoites (in) attached to the surface of monkey red cells. All sectionsexcept C are labelled with anti-band 3 antibodies, followed by gold-labelled protein A. A. Merozoite at an early stage ofattachment; note absence of label in the junction zone. Inset: an attached merozoite showing invagination of the red cellmembrane and formation of an internal vacuole (i>). B. TWO attached merozoites on a single red cell with internal vacuole. As inA inset, all regions of the red cell in contact with the merozoite are devoid of gold label, as are the membranes surrounding theinternal vacuoles (v). C. Laminar body in merozoite, apparently discharging electron-dense material into host cell vacuole (v).Bars, 1 /Jm.

Membrane proteins in parasitised red cell 697

Loyter et al. 1977); we have obtained evidence thatphosphorylation state may also be linked to the malarialinvasion process (Rangachari et al. 1986), so that aparallel with endocytosis clearly suggests itself. Wecannot of course eliminate other mechanisms, such asintroduction of parasite-derived membrane material,which sweeps away the host cell membrane proteins, orindeed selective proteolytic destruction of all epitopes forour antibodies.

Band 3 is missing from the area of host cell membranein contact with the parasite, from the intracellularvesicles and also from the membrane in the cavity, thatwe take to represent an incipient internal vesicle. Thisagrees with the observation of Aikawa et al. (1981), whofound that all internal vesicle membranes were largelydevoid of intramembrane particles. We have observed noincreased concentration of band 3 at the boundary of thejunction between the parasite and host cell, at whichthere is high electron density (Aikawa et al. 1978). Theabsence of host cell protein from the parasitophorousvacuole cannot in itself be regarded as strong evidence forthe view (Bannister^ al. 1976; Atkinson et al. 1987) thatthe parasitophorous vacuole membrane is not derivedfrom the red cell. To resolve this question it will benecessary to examine the membrane lipids. It should benoted that, according to Eisen (1977), the internalized P.chabaudi parasite in mouse red cells has a membrane thatstains strongly with anti-spectrin in immunofluor-escence. This would certainly signify that in this case theparasitophorous vacuole membrane is derived from thehost cell.

The membrane in the cavity in the red cell formed atthe attachment site remains intact (Fig. 4C) and thiscould be taken to argue that, in this system at least, theparasite may indeed surround itself with protein-depletedhost cell membrane. Whether the withdrawal of themembrane opposite the parasite is due to initiation of anendocytosis-like process, possibly requiring energy fromATP (Penniston et al. 1979), or is caused by theintroduction into the membrane of a destabilizing agent,secreted by the parasite, is not clear. Fig. 4C suggeststhat a transfer of material from the parasite does occur.This may be derived from the substance of the multila-mellar body seen in the parasite. The latter is notidentical with the 'whorls' seen in the rhoptries ofmerozoites (Bannister et al. 1986; Stewart et al. 1986), orindeed multilamellar liposomes (Bangham et al. 1965),both of which are characterized by lamellar spacings ofthe order of 5 nm. This is lower by an order of magnitudethan the spacing in the body seen in Fig. 4C, whichsuggests that it is not simply a lipid- or detergent-likematerial, and presumably contains larger molecularspecies, such as proteins.

We are indebted to Dr M. J. A. Tanner and to Dr A. J. Bainesfor generous gifts of anti-band 3 and anti-ankynn antibodies,and to J. M. Hopkins for help and advice on electron mi-croscopy. This work was supported by the UNDP/WorldBank/World Health Organization Special Programme for Re-search and Training in Tropical Diseases.

References

AIKAWA, M., MILLER, L. H., JOHNSON, J. & RABBEGE, J. (1978).

Erythrocyte entry by malaria parasites. A moving junction betweenerythrocyte and parasite. J. Cell Biol. 77, 72-82.

AIKAWA, M., MILLER, L. H., RABBEGE, J. & EPSTEIN, N. (1981).

Freeze-fracture study on the erythrocyte membrane dunng malanalparasite invasion. J. Cell Biol. 91, 55-62.

ATKINSON, C. T., AIKAWA, M., PERRY, G., FUJINO, T., BENNETT,

V., DAVIDSON, E. A. & HOWARD, R. J. (1987). Ultrastructurallocalization of erythrocyte cytoskeletal and integral membraneproteins in Plasmodium falcipantm-mfected erythrocytes. Eur.J. Cell Biol. 45, 192-199.

BANGHAM, A. D., STANDISH, M. M. & WATKJNS, J. C. (1965).

Diffusion of univalent ions across lamellae of swollenphospholipids.7. molec. Biol. 13, 238-252.

BANNISTER, L. H., MITCHELL, G. H., BUTCHER, G. A. & DENNIS, E.

G. (1986). Lamellar membranes associated with rhoptriea inerythrocytic merozoites of Plasmodium knowtesi: a clue to themechanism of invasion. Parasitology 92, 291-303.

BENNETT, V. & STENBUCK, P. J. (1980). Human erythrocyte ankyrin.Purification and properties. J. biol. Chem. 255, 2540-2548.

BROWN, G. V., CULVENOR, J. G., CREWTHER, P. E., BIANCO, A. E.,

COPPEL, R. L., SAINT, R. B., STAHL, H. D., KEMP, D. J. &

ANDERS, R. F. (1985). Localization of the ring-infected erythrocytesurface antigen (RESA) of Plasmodium falciparum in merozoitesand ring-infected erythrocytes. J. exp. Med. 162, 774-779.

DLUZEWSKI, A. R., FRYER, P. B., GRIFFITHS, S., RANGACHARI, K.,

WILSON, R. J. M. & GRATZER, W. B. (1988). Exclusion of red cellmembrane cytoskeleton from the parasitophorous vacuolemembrane of the internalised malaria parasite. Cell Biol. Int. Rep.12, 149.

DLUZEWSKI, A. R., LING, I. T., RANGACHARI, K., BATES, P. A. &

WILSON, R. J. M. (1984). A simple method for isolating viablemature parasites of Plasmodium falciparum from cultures. Trans.R. Soc. trap. Med. Hyg. 78, 622-624.

EISEN, H. (1977). Purification of intracellular forms of Plasmodiumchabaudi and their interactions with the erythrocyte membraneand with human serum. Bull. WHO 55, 333-338.

FOXWELL, B. M. J. & TANNER, M. J. A. (1981). Synthesis of theerythrocyte anion-transport protein. Immunochemical study of theincorporation into the plasma membrane of erythroid cells.Biochem.jf. 195, 129-137.

GAZITH, Y., LOYTER, Y., REICHLER, Y. & OHAD, I. (1976).

Correlation between changes in the membrane organization andsusceptibility to phospholipase C attack induced by ATP depletionof rat erythrocytes. Biochim. biophys. Ada 419, 479-492.

GRATZER, W. B. (1982). Preparation of spectrin. Meth. Enzvm. 85,475-480.

HARDY, B., BENSCH, K. G. & SCHRIER, S. L. (1979). Spectrin

rearrangement in erythrocyte ghost endocytosis. J. biol. Chem. 82,654-663.

Jl, T. H. & NlCOLSON, G. L. (1974). Lectin binding andperturbation of the outer surface of the cell membrane induces atransmembrane organizational alteration at the inner surface. Proc.natn. Acad. Sci. U.SA. 71, 2212-2216.

LAMBROS, C. & VANDENBERG, J. P. (1979). Synchronization ofPlasmodium falciparum intraerythrocytic stages in culture. J.Parasit. 65, 418-420.

LOYTER, A., BEN-ZAQUEN, R., MARASH, R. & MILNER, Y. (1977).Dephosphorylation of human erythrocytic membranes induced bySendai virus. Biochemistry 16, 3903—39.

M C L A K N , D. J., BANNISTER, L. H., TRIGG, P. I. & BUTCHER, G.

A. (1979). Freeze fracture studies of the interaction between themalaria parasite and the host erythrocyte in Plasmodium knowlesiinfections. Parasitology 79, 125-139.

NICOLSON, G. L. & PAINTER, R. G. (1973). Anionic sites of humanerythrocyte membranes. II. Antispectrin-induced transmembraneaggregation of the binding sites for positively charged colloidalparticles. J. Cell Biol. 59, 395-406.

PENNISTON, J. T., VAUGHAN, L. & NAKAMURA, M. (1979).

Endocytosis in erythrocytes and ghosts: occurrence at 0° after ATPpreincubation. Archs Biochem. biophys. 198, 339-348.

RANGACHARI, K., DLUZEWSKI, A. R., WILSON, R. J. M. & GRATZER,

698 A. R. Dluzewski et al.

W. B. (1986). Control of malarial invasion by phosphorylation of M. (1979). In vitro production and partial purification ofthe host cell membrane cytoskeleton. Nature, Land. 324, 364-365. Plasmodium falciparum antigen. Bull. WHO 57 (Suppl. 1), 75-82.

ROTH, J. (1982). The protein A-gold (pAg) technique. A qualitative STEWART, M. J., SCHULMAN, S. & VANDENBERG, J. P. (1986).and quantitative approach for antigen location in thin sections in Rhoptry secretion of membranous whorls by PlasmodiumTechniques in Immunocytochemistry, vol. 1 (ed. G. R. Bullock & falciparum merozoites. Am. J. trop. Med. Hyg. 35, 37-44.D. Petrusz), pp. 108-133. Academic Press, London. TRAGER, W. & JENSEN, J. B. (1976). Human malaria parasites in

ROTH, J., BENDAYAN, M., CARLEMALM, E., VILLIGER, W. & continuous culture. Science 193, 673-675.GARAUITO, M. (1981). Enhancement of structural preservation andimmunocytochemical staining in low temperature-embeddedpancreatic tissue. J. Histochem. Cytochem. 29, 663-671.

SIDDIQUI, W. A., KAN, S. C , KRONER, K. & RICHMOND-CRUM, S. (Received 16 November 1988-Accepted 12 January 1989)

Membrane proteins in parasitized red cell 699