EXPERIMENTAL PARASITOLOGY 13,62-72
(191)
Plasmodium falciparum: Analysis of the Interaction of Antigen Pfl!%/RESA with the Erythrocyte Membrane WIPAPORN RUANGJIRACHUPORN,* DETLEV DRENCKHAHN,~ *Department
RACHANEE UDOMSANGPETCH,” PETER PERLMANN* AND KLAVS
JAN CARLSON,* BERZINS*
of Immunology, Stockholm University, S-106 91 Stockholm, Sweden: and fDepartment Anatomy and Cell Biology, University of Marburg, D-3550 Marburg, Germany
of
RUANGJIRACHUPORN, W., UDOMSANGPETCH, R., CARLSSON, J., DRENCKHAHN, D., PERLMANN, P., AND BERZINS, K. (1991). Plasmodium fulciparum: Analysis of the interaction of antigen PflSYRESA with the erythrocyte membrane. Experimental Parasitology 73, 62-72. The location of the Plasmodium falciparum vaccine candidate antigen Pfl5YRESA
in the membrane of infected erythrocytes was analyzed by means of selective surface radioiodination and immunofluorescence of surface-modified cells. The lack of radiolabel in Pfl55/RESA as well as its localization by immunofluorescence similar to that of the Nterminal region of erythrocyte band 3 suggests that the antigen is associated with the cytoplasmic phase of the erythrocyte membrane. In concordance with this, PflSS/RESA was detected by immunofluorescence on the surface of inside out membrane vesicles from P. falciparum-infected erythrocytes. PflSYRESA from spent culture medium also bound to inside out membrane vesicles of normal erythrocytes as well as to cytoskeletal shells of such vesicles, but failed to bind to sealed right-side out membrane vesicles. Depletion of spectrin from the vesicles abolished antigen binding, suggesting that Pfl5YRESA association with the erythrocyte cytoskeleton is mediated by spectrin. 0 1’991AcademicPress.Inc. INDEX DESCRIPTORS AND ABBREVIATIONS: Plasmodium falciparum; Malaria, human; Protozoa, parasitic; Erythrocyte membrane vesicles; Cytoskeleton; Immunofluorescence; Erythrocyte membrane immunofluorescence (EMIF); Kilodalton (kDa); Monoclonal antibody (MAb); Phosphate-buffered saline (PBS); Polyacrylamide gel electrophoresis (PAGE); Relative molecular weight (M,); Ring stage-infected erythrocyte surface antigen (RESA); Sodium dodecyl sulfate (SDS).
et al. 1990). Pfl5YRESA appears to be in-
INTRODUCTION
PflSSIRESA is a Plasmodium falciparum antigen of M,. 155kDa previously thought to be associated with the micronemes and rhoptries of merozoites (Brown et al. 1985; Aikawa 1986; Uni et al. 1987). However, with improved analysis the antigen has now been localized to dense granules in the apical end of the merozoites (Aikawa et al. 1990). During or shortly after invasion the antigen is probably released into the parasitophorous vacuole space from where it is translocated by an unknown process (Aikawa et al. 1990) to become associated with the membrane of newly infected erythrocytes (Perlmann et al. 1984; Aikawa 1986; Uni et al. 1987) by interaction with the cytoskeleton (Brown et al. 1985; Foley 62 0014-4894/91
$3.00
Copyright0 1991by AcademicPress,Inc. All rights of reproductionin any form reserved.
volved in the invasion process as antibodies to the antigen inhibit invasion with high efficiency (Wshlin et al. 1984; Berzins et al. 1986; Perlmann et al. 1989; Ruangjirachuporn et al. 1988). The function of the antigen is, however, not known, but it has been suggested that its interaction with the erythrocyte cytoskeleton facilitates merozoite invasion (Uni et al. 1987). A homology between a repeated sequence of Pf155/ RESA and the N-terminal sequences of human erythrocyte band 3 also suggests functional homologies between the two proteins with regard to interactions with cytoskeletal and cytoplasmic erythrocyte proteins (Favaloro et al. 1986; Anders et al. 1987). Detection of Pfl55/RESA by immunofluorescence at the surface of P. falciparum-
P. falciparum: ANALYSIS0F ANTIGEN Pfl55l~Es~ infected erythrocytes requires a moditication of the erythrocyte membrane by treatment with fixative and air drying (Perlmann et al. 1984; Brown et al. 1985; Uni et al. 1987). Antibodies active in this modified immunofluorescence test (EMIF) are mainly directed against the C-terminal repeat region of PflWRESA. Thus, this part of the antigen is not exposed on the surface of intact erythrocytes to such a degree that it is accessible to antibodies or proteolytic enzymes (Perlmann et al. 1984; Brown et al. 1985; Uni et al. 1987). In this study we have analyzed the location and possible associations of Pf155/ RESA in the erythrocyte membrane. This was done with antibodies to Pfl55/RESA or to the band 3 protein by means of selective surface radiolabeling as well as immunofluorescence analyses of infected erythrocytes modified in various ways. In addition, the binding of Pfl55/RESA to different types of erythrocyte membrane vesicles was investigated. MATERIAL
AND METHODS
Parashes. P. falciparum of the Tanzanian strain F32 and the Ugandan strain Palo Alto were obtained from in vifro cultures maintained as described by Trager and Jensen (1976). Synchronization of the parasites was accomplished by Percoll (Pharmacia, Uppsala, Sweden) centrifugation according to Stanley et al. (1982). lodination. Surface labeling of P. falciparuminfected erythrocytes with ‘*‘I was accomplished by the lactoperoxidase-glucose oxidase method of Schenkein et al. (1972). The reaction was allowed to go on for 15 min and the erythrocytes were then washed five times in phosphate-buffered saline (PBS) and subsequently extracted with Triton X-100. SDS-PAGE and immunoblotting. Separation by SDS-polyacrylamide gel electrophoresis (SDSPAGE) was performed under reducing or nonreducing conditions in 5-15% linear gradient gels according to the method of Neville (1971) as modified by Cohen et al. (1977). Immunoblotting of the separated material was performed essentially as previously described (Batteiger et al. 1982; Berzins er al. 1983). Autoradiography was performed by exposure of dried gels or nitrocellulose to Fuiji RX film (Fuiji Co., Tokyo, Japan) at room temperature. Antibodies. Antibodies reactive with PflWRESA
63
were: the human monoclonal antibody (MAb) 3362 (Udomsangpetch er al. 1986), the mouse MAb 1Fl (Ruangjirachuporn et al. 1988), and rabbit antisera against synthetic peptides corresponding to repeated sequences of PflWRESA (Berzins et al. 1986; Carlsson et al. 1990).Antibodies reactive with human erythrocyte band 3 were: the mouse MAb 4C8F6 to the cytoplasmic N-terminal part, the mouse MAb 6FlC2 to a yet undetermined portion of the cytoplasmic domains, the mouse MAb 4GlOH9 to the cytoplasmic domain of the C-terminus, rabbit antiserum 11/86 to the cytoplasmic N-terminal domain and rabbit antiserum 28/84 directed against the whole protein. Fixation and treatment suspension. P. falciparum
of infected
erythrocytes
in
cultures of 5-10% parasitemia, containing preferentially early stages of the parasite, were washed in PBS and then fixed in 0.5% glutaraldehyde in PBS for 5 min. After washing in PBS the cells were divided for treatment at room temperature with either 0.1% (w/v) saponin in PBS for 30 min or 10% (w/v) Triton X-100 in PBS for 15 min. The cells were then stored in PBS at 4°C. Immunofluorescence. Erythrocyte membrane immunofluorescence (EMIF) was performed on glutaraldehyde fixed and air dried monolayers of P. falciparum infected erythrocytes as previously described (Perlmann et al. 1984). Ca*+, Mg’+, and ionophore A23187 treatment. Treatment of infected erythrocytes with Ca*+ or Mg* + together with the ionophore A23187 was performed as described by Lorand et al. (1987) using a concentration of 20 l&f A23187 and 1.5 mM Ca*+ or Mg’+. The incubation periods were 3 or 6 hr at 37°C. Erythrocyre membrane vesicles. Sealed right-side out vesicles or inside out vesicles were prepared by the method of Steck et al. (1970) from either normal human erythrocytes (bloodgroup 0, Rh+) or erythrocytes from P. fulciparum in vitro cultures with lO-15% parasitemia. The sidedness of the vesicles was assessed by immunofluorescence using antibodies to spectrin and the cytoplasmically located N-terminal part of band 3. In general, at least 90% of the vesicles showed the proper sidedness. Spectrin depletion of inside out vesicles by hypotonic treatment and EDTA as well as further KI extraction of these vesicles was performed as described by Bennett (1983). Cytoskeletal shells of inside out membrane vesicles were obtained by treatment with 1% Triton X-100 and further extraction with 0.5 M KC1 was performed as described by Sheetz (1979). Trypsin digestion of inside out membrane vesicles was done by incubation for 30 min at room temperature in 20 vol of trypsin in PBS (200 t&ml), followed by several washings in PBS containing soybean trypsin inhibitor (200 &ml) and phenylmethylsulphonyl fluoride (20 &ml). Binding of PflSSIRESA to erythrocyte membrane vesicles. Pelleted membrane vesicles (0.5 mg protein)
64
RUANGJIRACHUPORN
were suspended in 1 ml of spent medium from P. falcultures, prepared as previously described (Perlmann et al. 1984). After incubation for 30 min at room temperature the vesicles were washed by repeated centrifugation in PBS. Binding of PflWRESA to the vesicles was assessed by immunoblotting after separation of SO-pg samples by SDS-PAGE using the PflWRESA reactive MAbs lF1 or 33G2 for probing. In order to estimate the relative amounts of different erythrocyte membrane or cytoskeletal proteins associated with the various vesicles, duplicate samples were separated by SDS-PAGE, stained with Coomassie blue (Fairbanks et al. 1971), and then analyzed by scanning in a densitometer (Ultroscan laser LKB 2202, Bromma, Sweden). ciparum
RESULTS
In order to see if some part of Pfl55/ RESA is exposed on the surface of intact infected erythrocytes, surface radiolabeling was performed by lactoperoxidase catalyzed iodination. The labeling was done on P. fulciparum cultures where high parasitemia of ring-infected erythrocytes was obtained by Percoll gradient enrichment of late-stage-infected erythrocytes which then were allowed to develop and to reinvade new erythrocytes. The labeled erythrocytes were extracted with Triton X-100. Soluble and insoluble materials were separated by SDS-PAGE under reducing or nonreducing conditions and transferred to nitrocellulose. Autoradiography showed that labeling was restricted to the erythrocyte surface as intracellularly located proteins like spectrin (doublet of 240 and 220 kDa) and hemoglobin were not labeled (Fig. la). Furthermore, no labeling was seen in the M, 155 kDa region (Fig. la). Immunoblotting with the Pfl55/RESA reactive human monoclonal antibody 3362 showed that this antigen was present in the Triton X-100 insoluble material (Fig. lb). Immunoprecipitation experiments confirmed the absence of labeling of Pfl551RESA although the antigen was highly enriched in the immunoprecipitates as detected by subsequent immunoblotting (not shown). The association of PflSS/RESA with the erythrocyte membrane was further studied
ETAL.
by immunofluorescence using a panel of antibodies to the C-terminal repeat region of this antigen. It was previously shown that Pfl55/RESA is not accessible for reaction with antibodies on the surface of intact P. fulciparum-infected erythrocytes unless the membrane is modified, e.g., by glutaraldehyde fixation and air drying (Perlmann et al. 1984). Infected erythrocytes were glutaraldehyde-fixed in suspension and then treated in various ways to perturb the erythrocyte membrane. All Pfl55/RESA reactive antibodies showed the same pattern of reactivity in immunofluorescence with the different erythrocyte preparations as shown in Table I for the antibodies 3362 and 1Fl . The antibodies which gave strong erythrocyte membrane immunofluorescence of glutaraldehyde-fixed and airdried-infected erythrocytes also gave such staining, although weaker, of erythrocytes glutaraldehyde-fixed in suspension without subsequent air-drying (Fig. 2a). Treatment of glutaraldehyde-fixed erythrocytes with saponin or Triton X-100 exposed Pf155/ RESA to a higher degree as judged from immunofluorescence which was of similar strength as that obtained with air-dried preparations in the regular EMIF (Figs. 2b and 2~). In parallel, the different preparations were probed with antibodies to human erythrocyte band 3. The monoclonal antibodies 4CSF6, 6FlC2, and 461049 specific for different cytoplasmically located epitopes of the protein only reacted with glutaraldehyde-fixed erythrocytes after airdrying or Triton X-100 treatment (see Table 1 for 4CSF6 as example). In contrast, polyclonal rabbit antibodies to the cytoplasmic N-terminal part of band 3 showed a reactivity pattern similar to that of antiPfl55/RESA antibodies. However, while the latter antibodies only stained the membrane of infected erythrocytes, the antiband 3 antibodies stained both noninfected and infected cells. As expected, antibodies to the whole band 3 protein gave a strong surface immunofluorescence staining on all
P. fafciparum:
ANALYSIS
a
OF ANTIGEN
Pfl55IRESA
65
b
200 w
116 93
31
FIG. I. Autoradiogram (a) and immunoblotting (b) of SDS-PAGE run under reducing (lanes 1 and 4) or nonreducing conditions (lanes 2 and 3). Triton X-100 insoluble (lanes 1 and 2) and soluble components (lanes 3 and 4) from “‘1 surface-labeled P. fulciparum-infected erythrocytes were analyzed. The MAb 33Ci2 was used for probing in the immunoblotting. Numbers indicate approximate molecular weights in kDa. Arrows indicate Pfl5YRESA.
erythrocytes in all preparations, including intact untreated erythrocytes (Table I). In order to elucidate how PflSYRESA is associated with the erythrocyte membrane, ionophore A23187 was used. Treatment of human erythrocytes with Ca*+, but not Mg’+, in the presence of this ionophore causes the formation of membrane protein polymers, mainly by cross-linking of band 3, ankyrin, spectrin, and band 4.1 (Lorand et al. 1987). We submitted P. falciparum ring-infected erythrocytes to such treatment, followed by analysis in EMIF. Probing with antibodies to PflSYRESA gave a
membrane immunofluorescence of infected cells which was weaker than that seen with untreated cells or cells treated with ionophore and Mg’+ (Figs. 3a and 3b). The smooth staining of untreated erythrocytes or of erythrocytes treated with ionophore and Mg*+ when probed with antibodies to human erythrocyte band 3 was, however, turned into a patchy staining of cells treated with Ca*+/ionophore (Figs. 3c and 3d). The location of Pfl5YRESA at the cytoplasmic side of the erythrocyte membrane was further demonstrated by immunofluorescence of membrane vesicles prepared
RUANGJIRACHUPORN
ET AL.
TABLE I EMIF Analysis of P. falciparum-Infected Erythrocytes after Glutaraldehyde Fixation in Suspension and Various Treatments Treatment of erythrocytes Glutaraldehyde fixed Antibodies
Reactivity
Air-dried
None
Saponin
Triton
None
33G2 1Fl 4C8F6 11186 28184
PflSSIRESA Pfl WRESA Band 3 Band 3 Band 3
++ ++ ++* ++* ++*
+ + +* ++*
++ ++ ++* ++*
++ ++ ++* ++* ++*
++*
Notes. 33G2: human Mab, 1Fl: mouse Mab, 4C8F6: mouse Mab to the cytoplasmic N-terminal part, 11186: polyclonal rabbit antiserum to the cytoplasmic N-terminal domain, 28/84: polyclonal rabbit antiserum directed against the whole protein. + + : Strong immunofluorescence, + : weak immunofluorescence, - : negative; *: both infected and noninfected erythrocytes were stained.
from P. falciparum-infected erythrocytes. While sealed right-side out vesicles were not stained with antibodies to PflSYRESA, inside-out vesicles were strongly stained (Fig 4). The interaction of Pfl5YRESA with the cytoplasmic side of the erythrocyte membrane was also demonstrated in binding studies of the antigen from spent culture medium to membrane vesicles (Table II, Fig. 5). Pfl5YRESA did not bind to intact erythrocytes nor to sealed right-side out vesicles, but bound well to inside out vesicles and leaky right-side out vesicles. It also bound to cytoskeletal shells obtained by solubilization of the erythrocyte membrane with Triton X-100. This binding was intact after further extraction of the cytoskeletal shells with 0.5 M KC1 in 1% Triton X-100, a treatment removing most cytoskeletal proteins, leaving mainly spectrin and some actin (Table II, Fig. 5). Hypotonic treatment in the presence of EDTA of inside out vesicles gave an incomplete depletion of spectrin and resulted in a marked reduction of Pfl55/RESA binding. When such vesicles were completely depleted of spectrin by further extraction with 1 M KI, Pfl55/RESA binding was also completely abolished (Table II, Fig. 5). The protein nature of the erythrocyte components involved in Pfl55/RESA binding was indi-
cated by the finding that trypsin or chymotrypsin digestion of inside out vesicles abolished binding. Binding of Pfl55/RESA to the erythrocyte membrane was not dependent on divalent cations as it was not affected by chelating agents like EDTA or EGTA at 10 mM concentration. Furthermore, it was intact at high (1 M) concentrations of NaCl or KCl. Pfl55/RESA seemed not to need to be in a native conformation for binding as antigen from boiled culture supernatants or eluted from gels after SDS-PAGE bound to erythrocyte membrane vesicles, although giving slightly weaker signals in the immunoblotting assay. The repeat regions of Pfl55/RESA seemed not to be involved in its binding to the erythrocyte membrane as the corresponding synthetic peptides K(EENVEHDA),, K(EENV),, and K(DDEHVEEPTVA), did not affect binding at concentrations up to 200 p&ml. Antigen binding was similarly intact in the presence of Pfl55/RESA reactive rabbit antibodies to these peptides as well as that of the MAbs 1Fl or 3362. DISCUSSION
This study confirms and extends earlier
P. falciparum:
ANALYSIS
OF ANTIGEN
F’fl55lRESA
67
FIG. 2. Indirect immunofluorescence using P. fulciparum-infected erythrocytes glutaraldehydefixed in suspension only (a) or subsequently treated with 0.1% saponin (b) or 10% Triton X-100 (c, d). Probing was done with the human MAb 3362 (a-c) or the mouse Mab 4C8F6 directed against a cytoplasmic epitope of band 3 (d). The samples were counterstained with ethidium bromide to visualize the intracellular parasites.
results indicating that PflWRESA is not exposed on the surface of P. falciparuminfected erythrocytes (Perlmann et al. 1984; Brown et al. 1985; Uni et al. 1987). Previously it was found that the C-terminal repeat region of Pf155/RESA is inaccessible on intact erythrocytes for antibody binding (Perlmann et al. 1984; Uni et al. 1987)or for digestion with proteolytic enzymes (Perlmann et al. 1984). In addition, we have shown that this antigen is not labeled by selective lactoperoxidase-catalyzed radioiodination of the surface proteins of in-
fected erythrocytes. Thus, neither the immunodominant antigenic epitopes within the C-terminal repeat region of Pf155/ RESA nor its tyrosine residues (Favaloro et al. 1986) are present on the outside of the erythrocyte membrane. Furthermore, analysis by immunofluorescence of surfacemodified infected erythrocytes indicated that Pf155/RESA has a location similar to that of the cytoplasmic part of erythrocyte band 3. These data are in concordance with the view that PflSS/RESA is associated with the erythrocyte cytoskeleton as sug-
68
RUANGJIRACHUPORN
ET AL.
cultures after treatment with ionophore A23187 and Ca” FIG. 3. EMIF analysis of P. falcipnrum (a and c) or Mg*+ (b and d). The samples were probed with the mouse MAb IF1 (a and b) or rabbit antiserum 11186 directed against band 3 (c and d). Counter staining was performed with ethidium bromide to visualize the intracellular parasites.
gested by the lack of solubility of the anti- (Wahlin et al. 1984; Berzins et al. 1986; gen in nonionic detergents (Brown ef al. Ruangjirachuporn et al. 1988; Perlmann et 1985). The association of Pfl5YRESA with al. 1989) suggests that it is involved in the cytoskeletal components at the cytoplasmic invasion process. This has come under face of the erythrocyte membrane is also doubt because of the recent identification indicated by our binding studies as well as of a variant of the FCR3 strain of P. falciby the ability of the antigen to become parum which does not express Pfl55/RESA phosphorylated when bound to inside out but, nevertheless, grows well in in vitro culerythrocyte membrane vesicles (Anders et ture (Cappai ef al. 1989). However, the caal. 1987; Foley et al. 1990). pacity of these Pfl55/RESA negative paraThe function of Pfl55/RESA is not sites to grow in vivo has not been estabknown, but the high efficiency of antibodies lished and no such variants have thus far reactive with this antigen to inhibit P. fuf- been detected in the isolates of >200 paciparum merozoite invasion in vitro tients (Perlmann et al. 1987; S. Thaithong,
P. fakiparum:
ANALYSIS OF ANTIGEN
69
Pfl55lRESA
homologies with parts of the N-terminal sequence of human band 3 (Favaloro et al. 1986; Anders et al. 1987), homologies also reflected by antigenic cross-reactivity between the proteins (Holmquist et al. 1988). As the intracellularly located N-terminus of band 3 is involved in intermolecular interactions with cytoplasmic and cytoskeletal proteins (Low 1986), it is possible that Pf155/RESA may exert some of these interactions as well and thereby interfere with some of the band 3 functions (Anders et al. 1987). Interestingly, the homologies with FIG. 4. Indirect immunofluorescence staining of in- the C-terminal repeats of Pfl55/RESA also side out vesicles of membranes from P. falciparuminclude the part of the N-terminal band 3 infected erythrocytes with MAb IFI. sequence which serves as binding site for the P. falciparum aldolase (Dobeli et al. unpublished). In any event, the involve1990), suggesting yet another way for ment of Pfl%/RESA in the process of nor- Pf155/RESA to interfere with band 3 funcmal invasion is strongly supported by the tion. An association with the erythrocyte findings that rabbit anti-peptide antibodies cytoskeleton similar to that of Pfl55/RESA which are highly specific for epitopes in the has also been suggested for a P. falciparum C-terminal repeat region of the molecule rhoptry protein of M, 110 kDa (Samare strong inhibitors of invasion of Pf155/ Yellowe et al. 1988). It is possible that RESA-positive parasites but not of that of PflSS/RESA and the M, 110 kDa protein the Pf155/RESA-negative variants (Wahlin form the electron dense material observed in electron microscopy associated with the et al., in preparation). Our binding data indicate that Pf155/ cytoskeleton of infected erythrocytes (Taylor et al. 1987). However, the mechanism RESA binds to spectrin in the erythrocyte cytoskeleton, an association also suggested for translocation of PflSS/RESA from the by Foley et al. (1990). Sequences in the C- dense granules in the merozoites to the interminal repeat region of Pf155/RESA have side of the erythrocyte membrane remains TABLE II Erythrocyte Membrane Polypeptides in Various Erythrocyte Membrane Vesicles, Determined by Densitometry after SDS-PAGE and Binding to the Vesicles of Pfl55/RESA, Determined by Immunoblotting Type of vesicle Right side out Inside out Triton KC&T&on Hypotonic EDTA 1 M KI
Erythrocyte membrane polypeptide 1
2
2.1
3
4.1
4.2
5
6
7
+ + +
+ + +
+
+ + +
(+‘, -
+ + + (+I -
+
A -
+ + + + -
+ + -
+ + -
(:I + +
(t, + +
(t) -
PflWRESA binding + + A -
Note. (+) Denotes presence of low amounts. The polypeptides are designated as defined by Steck (1974). 1, o-subunit of spectrin; 2, B-subunit of spectrin; 2.1, ankyrin; 3, band 3, the major anionexchange protein; 4.1, band 4.1; 4.2, band 4.2; 5, actin; 6, glyceraldehyde-3-phosphate dehydrogenase; 7, band 7.
70
RUANGJIRACHU
PO ‘RN ET AL.
ABCDEF
ABCDEF
zoo-155
11693-
66-
-5
45-
-6 31-7 1
II
FIG. 5. Binding of PflWRESA
to various erythrocyte membrane vesicle preparations. I. Polypeptide pattern of the membrane vesicles in SDS-PAGE as revealed by staining with Coomassie blue. II. Immunoblotting using MAb IF1 for probing. A, Cytoskeletal shells; B, sealed right-side out membrane vesicles; C, inside out membrane vesicles; D, hypotonically EDTA-treated inside out membrane vesicles; E, KI extracted membrane vesicles; F, KCI-Triton-extracted cytoskeletal shells. Numbers to the left indicate approximate molecular weights in kDa and numbers to the right indicate the positions of the erythrocyte polypeptides as designated by Steck (1974). For definition of the polypeptides, see Table II.
a puzzle. The primary structure of Pf15Y RESA does not show any typical membrane-spanning hydrophobic sequences (Favaloro et al. 1986) and, accordingly, the antigen behaves as a soluble antigen in phase partition extractions with Triton X114 (Bianco et al. 1987). In P. knowlesi, the dense granules, corresponding to those containing Pf1551 RESA in P. fufciparum, release their contents into the parasitophorous vacuole after the merozoites have entered erythrocytes (Torii et al. 1989). By analogy, Pf155/RESA might be accumulated in the parasitophorous vacuole during invasion and then move from this location to the erythrocyte membrane by an unknown process (Aikawa et al. 1990). By this route of movement, Pfl55/RESA would not be expected to be
exposed and easily accessible to antibodies shown to inhibit invasion very efficiently (Wahlin et al. 1984; Berzins et al. 1986; Ruangjirachuporn et al. 1988; Perlmann et al. 1989). However, the mechanism of this inhibition involving Pf155/RESA as target antigen is presently unknown. In EMIF, antibodies to PflSS/RESA may be seen to stain the surface of erythrocytes to which a merozoite has attached but before it has entered the cell (Perlmann et al. 1984). Moreover, the antigen is also found in easily demonstrable amounts in the medium of P. fulciparum cultures (Perlmann et al. 1984; Carlsson et al. 1990). Taken together, such findings imply that the pathway of antigen release from dense granules in P. falciparum may be different from that postulated for P. knowlesi (Torii et al. 1989),
P.
fakiparum:
ANALYSIS OF ANTIGEN
or, alternatively, that PflWRESA is also present at other locations, not as yet defined by electron microscopy. Further work is needed to resolve these questions. ACKNOWLEDGMENTS This work was supported by grants from the UNDP/ World Bank/WHO Special Programme for Research and Training in Tropical Diseases, the Swedish Medical Research Council, and the Rockefeller Foundation Great Neglected Diseases Network. The support from Kabi AB and SmithKline Beecham Biologicals is gratefully acknowledged.
REFERENCES AIKAWA, M. 1986. Localization of protective malarial antigens by immunoelectron microscopy. Memorius do Institute Oswald0 Cruz 81 Suppl. II, 159-166. AIKAWA, M., TORII, M., SJ~LANDER, A., BERZINS, K., PERLMANN, P., AND MILLER, L. H. 1990. PflSS/RESA antigen is localized in dense granules of Plasmodium falciparum merozoites. Experimental Parasitology 71, 326-329. ANDERS, R. F., MURRAY, L. J., THOMAS, L. M., DAVERN, K. M., BROWN, G. V., AND KEMP, D. J. 1987. Structure and function of candidate vaccine antigens in Plasmodium falciparum. Biochemical Society Symposia 53, 103-l 14. BATTEIGER, B., NEWHALL, V. W. J., AND JONES, R. B. 1982. The use of Tween 20 as a blocking agent in the immunological detection of proteins transferred to nitrocellulose membranes. Journal oflmmunological Methods 55, 297-307. BENNETT, V. 1983. Proteins involved in membranecytoskeleton association in human erythrocytes: Spectrin, ankyrin, and band 3. In “Methods in Enzymology” (S. Fleischer and B. Fleischer, Eds.), Vol. 96, pp. 313-324. Academic Press, San Diego. BERZINS, K., PERLMANN, H., WAHLIN, B., CARLSSON, J., WAHLGREN, M., UDOMSANGPETCH, R., BJ~RKMAN, A., PATARROYO, M. E., AND PERLMANN, P. 1986. Rabbit and human antibodies to a repeated amino acid sequence of a Plasmodium falciparum antigen, Pfl55, react with the native protein and inhibit merozoite invasion. Proceedings of the National Academy of Sciences USA 88, 1065-1069. BERZINS, K., WAHLGREN, M., AND PERLMANN, P. 1983. Studies on the specificity of anti-erythrocyte antibodies in the serum of patients with malaria. Clinical and Experimental Immunology 54,313-318. BIANCO, A. E., CULVENOR, J. G., COPPEL, R. L., CREWTHER, P. E., MCINTYRE, P., FAVALORO, J. M., BROWN, G. V., KEMP, D. J., AND ANDERS, R. F. 1987. Putative glycophorin-binding protein is
PflWRESA
71
secreted from schizonts of Plasmodium falciparum. Molecular and Biochemical Parasitology 23, 91102. BJ~RKMAN, A., HEDMAN, P., BROHULT, J., WILLcox, M., DIAMANT, I., PEHRSSON, P. O., ROMBO, L., AND BENGTSSON, E. 1985. Different malaria control activities in an area of Liberia. Effects on malariometric parameters. Annals of Tropical Medicine and Parasitology 79, 239-246. BROWN, G. V., CULVENOR, J. G., CREWTHER, P. E., BIANCO, A. E., COPPEL, R. L., SAINT, R. B., STAHL, H.-D., KEMP, D. J., AND ANDERS, R. F. 1985. Localization of the ring-infected erythrocyte surface antigen (RESA) of Plasmodium falciparum in merozoites and ring-infected erythrocytes. Journal of Experimental Medicine 162, 774-779. CAPPAI, R., VAN SCHRAVENDIJK, M.-R., ANDERS, R. F., PETERSON, M. G., THOMAS, L. M., CowMAN, A. F., AND KEMP, D. J. 1989. Expression of the RESA gene in Plasmodium falciparum isolate FCR3 is prevented by a subtelomeric deletion. Molecular and Cellular Biology 9, 3584-3587. CARLSSON, J., BERZINS, K., PERLMANN, H., AND PERLMANN, P. 1991. Studies on Pfl55ZRESA and other soluble antigens from in vitro cultured Plasmodium flaciparum. Parasitology Research 77, 2732. CARLSSON, J., UDOMSANGPETCH, R., WI\HLIN, B., AHLBORG, N., BERZINS, K., AND PERLMANN, P. 1990. Plasmodium falciparum: Differential parasite reactivity of antibodies to repeated sequences in the antigen Pfl5YRESA. Experimental Parasitology 71, 314-325. COHEN, R. S., BLOMBERG, F., BERZINS, K., AND SIEKEVITZ, P. 1977. The structure of postsynaptic densities isolated from dog cerebral cortex. I. Overall morphology and protein composition. Journal of Cell Biology 74, 181-203. D~BELI, H., TREZCIAK, A., GILLESSEN, D., MATILE, H., SRIVASTAVA, I. K., PERRIN, L. H., JAKOB, P. E., AND CERTA, U. 1990. Expression, puritication, biochemical characterization, and inhibition of recombinant Plasmodium falciparum aldolase. Molecular and Biochemical Parasitology 41, 259-268. FAIRBANKS, G., STECK, T. L., AND WALLACH, D. F. H. 1971. Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10, 2606-2617. FAVALORO, J. M., COPPEL, R. L., CORCORAN, L. M., FOOTE, S. J., BROWN, G. V., ANDERS, R. F., AND KEMP, D. J. 1986. Structure of the RESA gene of Plasmodium falciparum. Nucleic Acids Research
14, 826.5-8278. FOLEY, M., MURRAY, L. J., AND ANDERS, R. F. 1990. The ring-infected erythrocyte surface antigen protein of Plasmodium falcipnrum is phosphorylated
72
RUANGJIRACHUPORN
upon association with the host cell membrane. MOlecular and Biochemical HOLMQUIST, WIGZELL,
Parasitology
G., UDOMSANGPETCH, H., AND PERLMANN,
38, 69-76.
R., BERZINS, K., P. 1988. Plasmo-
dium chabaudi antigen PchlOS, Plasmodium falciparum antigen Pfl55, and erythrocyte band 3 share cross-reactive epitopes. Infection and Immunity 56,
1545-1550. LORAND, L., BARNES, HAWKINS, M., AND
N., BRUNER-LORAND, J. A., MICHALSKA, M. 1987. Inhibi-
tion of protein cross-linking in Ca*+-enriched human erythrocytes and activated platelets. Biochemistry 26, 308-313. Low, P. S. 1986. Structure and function of the cytoplasmic domain of band 3: Center of erythrocyte membrane-peripheral protein interactions. Biochimica et Biophysics Acta 864, 145-167. NEVILLE, D. M. 1971. Molecular weight determinations of proteindodecyl sulfate complexes by gel electrophoresis in a discontinuous buffer system. Journal
of Biological
Chemistry
246, 6328-6334.
PERLMANN, H., BERZINS, K., WAHLGREN, M., CARLSSON, J., BJ~RKMAN, A., PATARROYO, M. E., AND PERLMANN, P. 1984. Antibodies in malarial
sera to parasite antigens in the membrane of erythrocytes infected with early asexual stages of Plasmodium falciparum. Journal cine 159, 1686-1704.
of Experimenal
Medi-
antigenic diversity in Pfl55, a major parasite antigen in membranes of erythrocytes infected with PlasmoJournal
of Clinical
protein into the plasma membrane of host erythrocytes. Journal of Cell Biology 106, 1507-1513. SCHENKEIN, I., LEVY, M., AND UHR, J. W. 1972. The use of glucose oxidase as a generator of H,O, in the enzymic radioiodination of components of cell surfaces. Cellular Immunology 5, 49&493. SHEETZ, M. P. 1979. Integral membrane protein interaction with Triton cytoskeletons of erythrocytes. Biochimica et Biophysics Acta 557, 122-134. STANLEY, H. A., LANGRETH, S. G., REESE, R. T., AND
TRACER,
Microbiology
25, 2347-2354. PERLMANN, H., PERLMANN, P., BERZINS, K., WAHLIN, B., TROYE-BLOMBERG, M., HAGSTEDT, M., ANDERSSON, I., HUGH, B., PETERSEN, E., AND BJ~RKMAN, A. 1989. Dissection of the human anti-
body response to the malaria antigen Pfl55/RESA into epitope specific components. Immunological Reviews, 112, 115-132. RUANGJIRACHUPORN, W., WAHLIN, B., PERLMANN, H., CARLSSON, J., BERZINS, K., WAHLGREN, M., UDOMSANGPETCH, R., WIGZELL, H., AND PERLMANN, P. 1988. Monoclonal antibodies to a syn-
thetic peptide corresponding to a repeated sequence in the Plasmodium falciparum antigen Pfl55. Molecular and Biochemical Parasitology 29, 19-28. SAM-YELLOWE, T. Y., SHIO, H., AND PERKINS, M. 1988. Secretion of Plasmodium falciparum rhoptry
W.
1982.
Plasmodium
falciparum
merozoites: Isolation by density gradient centrifugation using Percoll and antigenic analysis. Journal of Parasitology 68, 1059-1067. STECK, T. L. 1974. The organization of proteins in the human red blood cell membrane. A review. Journal of Cell Biology 62, l-19. STECK, T. L., WEINSTEIN, R. S., STRAUS, J. H., AND WALLACH, D. F. H. 1970. Inside out red cell mem-
brane vesicles: Preparation and purification. Science 168, 255-257. TAYLOR, D. W., PARRA, M., AND STEARNS, M. E. 1987. Plasmodium falciparum: Fine structural changes in the cytoskeletons of infected erythrocytes. Experimental Parasitology 64, 178187. J. H., MILLER, L. H., AND TORII, M., ADAMS, AIKAWA, M. 1989.Release of merozoite dense granules during erythrocyte invasion by Plasmodium knowlesi.
PERLMANN, H., BERZINS, K., WAHLIN, B., UDOMSANGPETCH, R., RUANGJIRACHUPORN, W., WAHLGREN, M., AND PERLMANN, P. 1987. Absence of
dium falciparum.
ET AL.
Infection
and Immunity
51, 3238-3233.
J. B. 1976. Human malaria parasites in continuous culture. Science 193, 673675. UDOMSANGPETCH, R., LUNDGREN, K., BERZINS, K., PERLMANN, H., TROYE-BLOMBERG, M., CARLSSON,
TRAGER,
W., AND JENSEN,
J., WAHLGREN, M., PERLMANN, P., AND BJ~RKMAN, A. 1986. Human monoclonal antibodies to
Pfl55, a major antigen of malaria parasite PlasmoScience 231, 57-59. UNI, S., MASUDA, A., STEWART, M. J., IGARASHI, I., NUSSENZWEIG, R., AND AIKAWA, M. 1987. Ultrastructural localization of the 150/130kD antigens in sexual and asexual blood stages of Plasmodium falciparum-infected human erythrocytes. American Journal of Tropical Medicine and Hygiene 36, 481488. WAHLIN, B., WAHLGREN, M., PERLMANN, H., dium falciparum.
BERZINS, K., BJORKMAN, A., PATARROYO, M. E., AND PERLMANN, P. 1984. Human antibodies to a M,
155,000Plasmodium falciparum antigen efficiently inhibit merozoite invasion. Proceedings of the National Academy of Sciences USA 81, 7912-7916. Received 6 November 1990; accepted with revision 19 February 1991
(191)
Plasmodium falciparum: Analysis of the Interaction of Antigen Pfl!%/RESA with the Erythrocyte Membrane WIPAPORN RUANGJIRACHUPORN,* DETLEV DRENCKHAHN,~ *Department
RACHANEE UDOMSANGPETCH,” PETER PERLMANN* AND KLAVS
JAN CARLSON,* BERZINS*
of Immunology, Stockholm University, S-106 91 Stockholm, Sweden: and fDepartment Anatomy and Cell Biology, University of Marburg, D-3550 Marburg, Germany
of
RUANGJIRACHUPORN, W., UDOMSANGPETCH, R., CARLSSON, J., DRENCKHAHN, D., PERLMANN, P., AND BERZINS, K. (1991). Plasmodium fulciparum: Analysis of the interaction of antigen PflSYRESA with the erythrocyte membrane. Experimental Parasitology 73, 62-72. The location of the Plasmodium falciparum vaccine candidate antigen Pfl5YRESA
in the membrane of infected erythrocytes was analyzed by means of selective surface radioiodination and immunofluorescence of surface-modified cells. The lack of radiolabel in Pfl55/RESA as well as its localization by immunofluorescence similar to that of the Nterminal region of erythrocyte band 3 suggests that the antigen is associated with the cytoplasmic phase of the erythrocyte membrane. In concordance with this, PflSS/RESA was detected by immunofluorescence on the surface of inside out membrane vesicles from P. falciparum-infected erythrocytes. PflSYRESA from spent culture medium also bound to inside out membrane vesicles of normal erythrocytes as well as to cytoskeletal shells of such vesicles, but failed to bind to sealed right-side out membrane vesicles. Depletion of spectrin from the vesicles abolished antigen binding, suggesting that Pfl5YRESA association with the erythrocyte cytoskeleton is mediated by spectrin. 0 1’991AcademicPress.Inc. INDEX DESCRIPTORS AND ABBREVIATIONS: Plasmodium falciparum; Malaria, human; Protozoa, parasitic; Erythrocyte membrane vesicles; Cytoskeleton; Immunofluorescence; Erythrocyte membrane immunofluorescence (EMIF); Kilodalton (kDa); Monoclonal antibody (MAb); Phosphate-buffered saline (PBS); Polyacrylamide gel electrophoresis (PAGE); Relative molecular weight (M,); Ring stage-infected erythrocyte surface antigen (RESA); Sodium dodecyl sulfate (SDS).
et al. 1990). Pfl5YRESA appears to be in-
INTRODUCTION
PflSSIRESA is a Plasmodium falciparum antigen of M,. 155kDa previously thought to be associated with the micronemes and rhoptries of merozoites (Brown et al. 1985; Aikawa 1986; Uni et al. 1987). However, with improved analysis the antigen has now been localized to dense granules in the apical end of the merozoites (Aikawa et al. 1990). During or shortly after invasion the antigen is probably released into the parasitophorous vacuole space from where it is translocated by an unknown process (Aikawa et al. 1990) to become associated with the membrane of newly infected erythrocytes (Perlmann et al. 1984; Aikawa 1986; Uni et al. 1987) by interaction with the cytoskeleton (Brown et al. 1985; Foley 62 0014-4894/91
$3.00
Copyright0 1991by AcademicPress,Inc. All rights of reproductionin any form reserved.
volved in the invasion process as antibodies to the antigen inhibit invasion with high efficiency (Wshlin et al. 1984; Berzins et al. 1986; Perlmann et al. 1989; Ruangjirachuporn et al. 1988). The function of the antigen is, however, not known, but it has been suggested that its interaction with the erythrocyte cytoskeleton facilitates merozoite invasion (Uni et al. 1987). A homology between a repeated sequence of Pf155/ RESA and the N-terminal sequences of human erythrocyte band 3 also suggests functional homologies between the two proteins with regard to interactions with cytoskeletal and cytoplasmic erythrocyte proteins (Favaloro et al. 1986; Anders et al. 1987). Detection of Pfl55/RESA by immunofluorescence at the surface of P. falciparum-
P. falciparum: ANALYSIS0F ANTIGEN Pfl55l~Es~ infected erythrocytes requires a moditication of the erythrocyte membrane by treatment with fixative and air drying (Perlmann et al. 1984; Brown et al. 1985; Uni et al. 1987). Antibodies active in this modified immunofluorescence test (EMIF) are mainly directed against the C-terminal repeat region of PflWRESA. Thus, this part of the antigen is not exposed on the surface of intact erythrocytes to such a degree that it is accessible to antibodies or proteolytic enzymes (Perlmann et al. 1984; Brown et al. 1985; Uni et al. 1987). In this study we have analyzed the location and possible associations of Pf155/ RESA in the erythrocyte membrane. This was done with antibodies to Pfl55/RESA or to the band 3 protein by means of selective surface radiolabeling as well as immunofluorescence analyses of infected erythrocytes modified in various ways. In addition, the binding of Pfl55/RESA to different types of erythrocyte membrane vesicles was investigated. MATERIAL
AND METHODS
Parashes. P. falciparum of the Tanzanian strain F32 and the Ugandan strain Palo Alto were obtained from in vifro cultures maintained as described by Trager and Jensen (1976). Synchronization of the parasites was accomplished by Percoll (Pharmacia, Uppsala, Sweden) centrifugation according to Stanley et al. (1982). lodination. Surface labeling of P. falciparuminfected erythrocytes with ‘*‘I was accomplished by the lactoperoxidase-glucose oxidase method of Schenkein et al. (1972). The reaction was allowed to go on for 15 min and the erythrocytes were then washed five times in phosphate-buffered saline (PBS) and subsequently extracted with Triton X-100. SDS-PAGE and immunoblotting. Separation by SDS-polyacrylamide gel electrophoresis (SDSPAGE) was performed under reducing or nonreducing conditions in 5-15% linear gradient gels according to the method of Neville (1971) as modified by Cohen et al. (1977). Immunoblotting of the separated material was performed essentially as previously described (Batteiger et al. 1982; Berzins er al. 1983). Autoradiography was performed by exposure of dried gels or nitrocellulose to Fuiji RX film (Fuiji Co., Tokyo, Japan) at room temperature. Antibodies. Antibodies reactive with PflWRESA
63
were: the human monoclonal antibody (MAb) 3362 (Udomsangpetch er al. 1986), the mouse MAb 1Fl (Ruangjirachuporn et al. 1988), and rabbit antisera against synthetic peptides corresponding to repeated sequences of PflWRESA (Berzins et al. 1986; Carlsson et al. 1990).Antibodies reactive with human erythrocyte band 3 were: the mouse MAb 4C8F6 to the cytoplasmic N-terminal part, the mouse MAb 6FlC2 to a yet undetermined portion of the cytoplasmic domains, the mouse MAb 4GlOH9 to the cytoplasmic domain of the C-terminus, rabbit antiserum 11/86 to the cytoplasmic N-terminal domain and rabbit antiserum 28/84 directed against the whole protein. Fixation and treatment suspension. P. falciparum
of infected
erythrocytes
in
cultures of 5-10% parasitemia, containing preferentially early stages of the parasite, were washed in PBS and then fixed in 0.5% glutaraldehyde in PBS for 5 min. After washing in PBS the cells were divided for treatment at room temperature with either 0.1% (w/v) saponin in PBS for 30 min or 10% (w/v) Triton X-100 in PBS for 15 min. The cells were then stored in PBS at 4°C. Immunofluorescence. Erythrocyte membrane immunofluorescence (EMIF) was performed on glutaraldehyde fixed and air dried monolayers of P. falciparum infected erythrocytes as previously described (Perlmann et al. 1984). Ca*+, Mg’+, and ionophore A23187 treatment. Treatment of infected erythrocytes with Ca*+ or Mg* + together with the ionophore A23187 was performed as described by Lorand et al. (1987) using a concentration of 20 l&f A23187 and 1.5 mM Ca*+ or Mg’+. The incubation periods were 3 or 6 hr at 37°C. Erythrocyre membrane vesicles. Sealed right-side out vesicles or inside out vesicles were prepared by the method of Steck et al. (1970) from either normal human erythrocytes (bloodgroup 0, Rh+) or erythrocytes from P. fulciparum in vitro cultures with lO-15% parasitemia. The sidedness of the vesicles was assessed by immunofluorescence using antibodies to spectrin and the cytoplasmically located N-terminal part of band 3. In general, at least 90% of the vesicles showed the proper sidedness. Spectrin depletion of inside out vesicles by hypotonic treatment and EDTA as well as further KI extraction of these vesicles was performed as described by Bennett (1983). Cytoskeletal shells of inside out membrane vesicles were obtained by treatment with 1% Triton X-100 and further extraction with 0.5 M KC1 was performed as described by Sheetz (1979). Trypsin digestion of inside out membrane vesicles was done by incubation for 30 min at room temperature in 20 vol of trypsin in PBS (200 t&ml), followed by several washings in PBS containing soybean trypsin inhibitor (200 &ml) and phenylmethylsulphonyl fluoride (20 &ml). Binding of PflSSIRESA to erythrocyte membrane vesicles. Pelleted membrane vesicles (0.5 mg protein)
64
RUANGJIRACHUPORN
were suspended in 1 ml of spent medium from P. falcultures, prepared as previously described (Perlmann et al. 1984). After incubation for 30 min at room temperature the vesicles were washed by repeated centrifugation in PBS. Binding of PflWRESA to the vesicles was assessed by immunoblotting after separation of SO-pg samples by SDS-PAGE using the PflWRESA reactive MAbs lF1 or 33G2 for probing. In order to estimate the relative amounts of different erythrocyte membrane or cytoskeletal proteins associated with the various vesicles, duplicate samples were separated by SDS-PAGE, stained with Coomassie blue (Fairbanks et al. 1971), and then analyzed by scanning in a densitometer (Ultroscan laser LKB 2202, Bromma, Sweden). ciparum
RESULTS
In order to see if some part of Pfl55/ RESA is exposed on the surface of intact infected erythrocytes, surface radiolabeling was performed by lactoperoxidase catalyzed iodination. The labeling was done on P. fulciparum cultures where high parasitemia of ring-infected erythrocytes was obtained by Percoll gradient enrichment of late-stage-infected erythrocytes which then were allowed to develop and to reinvade new erythrocytes. The labeled erythrocytes were extracted with Triton X-100. Soluble and insoluble materials were separated by SDS-PAGE under reducing or nonreducing conditions and transferred to nitrocellulose. Autoradiography showed that labeling was restricted to the erythrocyte surface as intracellularly located proteins like spectrin (doublet of 240 and 220 kDa) and hemoglobin were not labeled (Fig. la). Furthermore, no labeling was seen in the M, 155 kDa region (Fig. la). Immunoblotting with the Pfl55/RESA reactive human monoclonal antibody 3362 showed that this antigen was present in the Triton X-100 insoluble material (Fig. lb). Immunoprecipitation experiments confirmed the absence of labeling of Pfl551RESA although the antigen was highly enriched in the immunoprecipitates as detected by subsequent immunoblotting (not shown). The association of PflSS/RESA with the erythrocyte membrane was further studied
ETAL.
by immunofluorescence using a panel of antibodies to the C-terminal repeat region of this antigen. It was previously shown that Pfl55/RESA is not accessible for reaction with antibodies on the surface of intact P. fulciparum-infected erythrocytes unless the membrane is modified, e.g., by glutaraldehyde fixation and air drying (Perlmann et al. 1984). Infected erythrocytes were glutaraldehyde-fixed in suspension and then treated in various ways to perturb the erythrocyte membrane. All Pfl55/RESA reactive antibodies showed the same pattern of reactivity in immunofluorescence with the different erythrocyte preparations as shown in Table I for the antibodies 3362 and 1Fl . The antibodies which gave strong erythrocyte membrane immunofluorescence of glutaraldehyde-fixed and airdried-infected erythrocytes also gave such staining, although weaker, of erythrocytes glutaraldehyde-fixed in suspension without subsequent air-drying (Fig. 2a). Treatment of glutaraldehyde-fixed erythrocytes with saponin or Triton X-100 exposed Pf155/ RESA to a higher degree as judged from immunofluorescence which was of similar strength as that obtained with air-dried preparations in the regular EMIF (Figs. 2b and 2~). In parallel, the different preparations were probed with antibodies to human erythrocyte band 3. The monoclonal antibodies 4CSF6, 6FlC2, and 461049 specific for different cytoplasmically located epitopes of the protein only reacted with glutaraldehyde-fixed erythrocytes after airdrying or Triton X-100 treatment (see Table 1 for 4CSF6 as example). In contrast, polyclonal rabbit antibodies to the cytoplasmic N-terminal part of band 3 showed a reactivity pattern similar to that of antiPfl55/RESA antibodies. However, while the latter antibodies only stained the membrane of infected erythrocytes, the antiband 3 antibodies stained both noninfected and infected cells. As expected, antibodies to the whole band 3 protein gave a strong surface immunofluorescence staining on all
P. fafciparum:
ANALYSIS
a
OF ANTIGEN
Pfl55IRESA
65
b
200 w
116 93
31
FIG. I. Autoradiogram (a) and immunoblotting (b) of SDS-PAGE run under reducing (lanes 1 and 4) or nonreducing conditions (lanes 2 and 3). Triton X-100 insoluble (lanes 1 and 2) and soluble components (lanes 3 and 4) from “‘1 surface-labeled P. fulciparum-infected erythrocytes were analyzed. The MAb 33Ci2 was used for probing in the immunoblotting. Numbers indicate approximate molecular weights in kDa. Arrows indicate Pfl5YRESA.
erythrocytes in all preparations, including intact untreated erythrocytes (Table I). In order to elucidate how PflSYRESA is associated with the erythrocyte membrane, ionophore A23187 was used. Treatment of human erythrocytes with Ca*+, but not Mg’+, in the presence of this ionophore causes the formation of membrane protein polymers, mainly by cross-linking of band 3, ankyrin, spectrin, and band 4.1 (Lorand et al. 1987). We submitted P. falciparum ring-infected erythrocytes to such treatment, followed by analysis in EMIF. Probing with antibodies to PflSYRESA gave a
membrane immunofluorescence of infected cells which was weaker than that seen with untreated cells or cells treated with ionophore and Mg’+ (Figs. 3a and 3b). The smooth staining of untreated erythrocytes or of erythrocytes treated with ionophore and Mg*+ when probed with antibodies to human erythrocyte band 3 was, however, turned into a patchy staining of cells treated with Ca*+/ionophore (Figs. 3c and 3d). The location of Pfl5YRESA at the cytoplasmic side of the erythrocyte membrane was further demonstrated by immunofluorescence of membrane vesicles prepared
RUANGJIRACHUPORN
ET AL.
TABLE I EMIF Analysis of P. falciparum-Infected Erythrocytes after Glutaraldehyde Fixation in Suspension and Various Treatments Treatment of erythrocytes Glutaraldehyde fixed Antibodies
Reactivity
Air-dried
None
Saponin
Triton
None
33G2 1Fl 4C8F6 11186 28184
PflSSIRESA Pfl WRESA Band 3 Band 3 Band 3
++ ++ ++* ++* ++*
+ + +* ++*
++ ++ ++* ++*
++ ++ ++* ++* ++*
++*
Notes. 33G2: human Mab, 1Fl: mouse Mab, 4C8F6: mouse Mab to the cytoplasmic N-terminal part, 11186: polyclonal rabbit antiserum to the cytoplasmic N-terminal domain, 28/84: polyclonal rabbit antiserum directed against the whole protein. + + : Strong immunofluorescence, + : weak immunofluorescence, - : negative; *: both infected and noninfected erythrocytes were stained.
from P. falciparum-infected erythrocytes. While sealed right-side out vesicles were not stained with antibodies to PflSYRESA, inside-out vesicles were strongly stained (Fig 4). The interaction of Pfl5YRESA with the cytoplasmic side of the erythrocyte membrane was also demonstrated in binding studies of the antigen from spent culture medium to membrane vesicles (Table II, Fig. 5). Pfl5YRESA did not bind to intact erythrocytes nor to sealed right-side out vesicles, but bound well to inside out vesicles and leaky right-side out vesicles. It also bound to cytoskeletal shells obtained by solubilization of the erythrocyte membrane with Triton X-100. This binding was intact after further extraction of the cytoskeletal shells with 0.5 M KC1 in 1% Triton X-100, a treatment removing most cytoskeletal proteins, leaving mainly spectrin and some actin (Table II, Fig. 5). Hypotonic treatment in the presence of EDTA of inside out vesicles gave an incomplete depletion of spectrin and resulted in a marked reduction of Pfl55/RESA binding. When such vesicles were completely depleted of spectrin by further extraction with 1 M KI, Pfl55/RESA binding was also completely abolished (Table II, Fig. 5). The protein nature of the erythrocyte components involved in Pfl55/RESA binding was indi-
cated by the finding that trypsin or chymotrypsin digestion of inside out vesicles abolished binding. Binding of Pfl55/RESA to the erythrocyte membrane was not dependent on divalent cations as it was not affected by chelating agents like EDTA or EGTA at 10 mM concentration. Furthermore, it was intact at high (1 M) concentrations of NaCl or KCl. Pfl55/RESA seemed not to need to be in a native conformation for binding as antigen from boiled culture supernatants or eluted from gels after SDS-PAGE bound to erythrocyte membrane vesicles, although giving slightly weaker signals in the immunoblotting assay. The repeat regions of Pfl55/RESA seemed not to be involved in its binding to the erythrocyte membrane as the corresponding synthetic peptides K(EENVEHDA),, K(EENV),, and K(DDEHVEEPTVA), did not affect binding at concentrations up to 200 p&ml. Antigen binding was similarly intact in the presence of Pfl55/RESA reactive rabbit antibodies to these peptides as well as that of the MAbs 1Fl or 3362. DISCUSSION
This study confirms and extends earlier
P. falciparum:
ANALYSIS
OF ANTIGEN
F’fl55lRESA
67
FIG. 2. Indirect immunofluorescence using P. fulciparum-infected erythrocytes glutaraldehydefixed in suspension only (a) or subsequently treated with 0.1% saponin (b) or 10% Triton X-100 (c, d). Probing was done with the human MAb 3362 (a-c) or the mouse Mab 4C8F6 directed against a cytoplasmic epitope of band 3 (d). The samples were counterstained with ethidium bromide to visualize the intracellular parasites.
results indicating that PflWRESA is not exposed on the surface of P. falciparuminfected erythrocytes (Perlmann et al. 1984; Brown et al. 1985; Uni et al. 1987). Previously it was found that the C-terminal repeat region of Pf155/RESA is inaccessible on intact erythrocytes for antibody binding (Perlmann et al. 1984; Uni et al. 1987)or for digestion with proteolytic enzymes (Perlmann et al. 1984). In addition, we have shown that this antigen is not labeled by selective lactoperoxidase-catalyzed radioiodination of the surface proteins of in-
fected erythrocytes. Thus, neither the immunodominant antigenic epitopes within the C-terminal repeat region of Pf155/ RESA nor its tyrosine residues (Favaloro et al. 1986) are present on the outside of the erythrocyte membrane. Furthermore, analysis by immunofluorescence of surfacemodified infected erythrocytes indicated that Pf155/RESA has a location similar to that of the cytoplasmic part of erythrocyte band 3. These data are in concordance with the view that PflSS/RESA is associated with the erythrocyte cytoskeleton as sug-
68
RUANGJIRACHUPORN
ET AL.
cultures after treatment with ionophore A23187 and Ca” FIG. 3. EMIF analysis of P. falcipnrum (a and c) or Mg*+ (b and d). The samples were probed with the mouse MAb IF1 (a and b) or rabbit antiserum 11186 directed against band 3 (c and d). Counter staining was performed with ethidium bromide to visualize the intracellular parasites.
gested by the lack of solubility of the anti- (Wahlin et al. 1984; Berzins et al. 1986; gen in nonionic detergents (Brown ef al. Ruangjirachuporn et al. 1988; Perlmann et 1985). The association of Pfl5YRESA with al. 1989) suggests that it is involved in the cytoskeletal components at the cytoplasmic invasion process. This has come under face of the erythrocyte membrane is also doubt because of the recent identification indicated by our binding studies as well as of a variant of the FCR3 strain of P. falciby the ability of the antigen to become parum which does not express Pfl55/RESA phosphorylated when bound to inside out but, nevertheless, grows well in in vitro culerythrocyte membrane vesicles (Anders et ture (Cappai ef al. 1989). However, the caal. 1987; Foley et al. 1990). pacity of these Pfl55/RESA negative paraThe function of Pfl55/RESA is not sites to grow in vivo has not been estabknown, but the high efficiency of antibodies lished and no such variants have thus far reactive with this antigen to inhibit P. fuf- been detected in the isolates of >200 paciparum merozoite invasion in vitro tients (Perlmann et al. 1987; S. Thaithong,
P. fakiparum:
ANALYSIS OF ANTIGEN
69
Pfl55lRESA
homologies with parts of the N-terminal sequence of human band 3 (Favaloro et al. 1986; Anders et al. 1987), homologies also reflected by antigenic cross-reactivity between the proteins (Holmquist et al. 1988). As the intracellularly located N-terminus of band 3 is involved in intermolecular interactions with cytoplasmic and cytoskeletal proteins (Low 1986), it is possible that Pf155/RESA may exert some of these interactions as well and thereby interfere with some of the band 3 functions (Anders et al. 1987). Interestingly, the homologies with FIG. 4. Indirect immunofluorescence staining of in- the C-terminal repeats of Pfl55/RESA also side out vesicles of membranes from P. falciparuminclude the part of the N-terminal band 3 infected erythrocytes with MAb IFI. sequence which serves as binding site for the P. falciparum aldolase (Dobeli et al. unpublished). In any event, the involve1990), suggesting yet another way for ment of Pfl%/RESA in the process of nor- Pf155/RESA to interfere with band 3 funcmal invasion is strongly supported by the tion. An association with the erythrocyte findings that rabbit anti-peptide antibodies cytoskeleton similar to that of Pfl55/RESA which are highly specific for epitopes in the has also been suggested for a P. falciparum C-terminal repeat region of the molecule rhoptry protein of M, 110 kDa (Samare strong inhibitors of invasion of Pf155/ Yellowe et al. 1988). It is possible that RESA-positive parasites but not of that of PflSS/RESA and the M, 110 kDa protein the Pf155/RESA-negative variants (Wahlin form the electron dense material observed in electron microscopy associated with the et al., in preparation). Our binding data indicate that Pf155/ cytoskeleton of infected erythrocytes (Taylor et al. 1987). However, the mechanism RESA binds to spectrin in the erythrocyte cytoskeleton, an association also suggested for translocation of PflSS/RESA from the by Foley et al. (1990). Sequences in the C- dense granules in the merozoites to the interminal repeat region of Pf155/RESA have side of the erythrocyte membrane remains TABLE II Erythrocyte Membrane Polypeptides in Various Erythrocyte Membrane Vesicles, Determined by Densitometry after SDS-PAGE and Binding to the Vesicles of Pfl55/RESA, Determined by Immunoblotting Type of vesicle Right side out Inside out Triton KC&T&on Hypotonic EDTA 1 M KI
Erythrocyte membrane polypeptide 1
2
2.1
3
4.1
4.2
5
6
7
+ + +
+ + +
+
+ + +
(+‘, -
+ + + (+I -
+
A -
+ + + + -
+ + -
+ + -
(:I + +
(t, + +
(t) -
PflWRESA binding + + A -
Note. (+) Denotes presence of low amounts. The polypeptides are designated as defined by Steck (1974). 1, o-subunit of spectrin; 2, B-subunit of spectrin; 2.1, ankyrin; 3, band 3, the major anionexchange protein; 4.1, band 4.1; 4.2, band 4.2; 5, actin; 6, glyceraldehyde-3-phosphate dehydrogenase; 7, band 7.
70
RUANGJIRACHU
PO ‘RN ET AL.
ABCDEF
ABCDEF
zoo-155
11693-
66-
-5
45-
-6 31-7 1
II
FIG. 5. Binding of PflWRESA
to various erythrocyte membrane vesicle preparations. I. Polypeptide pattern of the membrane vesicles in SDS-PAGE as revealed by staining with Coomassie blue. II. Immunoblotting using MAb IF1 for probing. A, Cytoskeletal shells; B, sealed right-side out membrane vesicles; C, inside out membrane vesicles; D, hypotonically EDTA-treated inside out membrane vesicles; E, KI extracted membrane vesicles; F, KCI-Triton-extracted cytoskeletal shells. Numbers to the left indicate approximate molecular weights in kDa and numbers to the right indicate the positions of the erythrocyte polypeptides as designated by Steck (1974). For definition of the polypeptides, see Table II.
a puzzle. The primary structure of Pf15Y RESA does not show any typical membrane-spanning hydrophobic sequences (Favaloro et al. 1986) and, accordingly, the antigen behaves as a soluble antigen in phase partition extractions with Triton X114 (Bianco et al. 1987). In P. knowlesi, the dense granules, corresponding to those containing Pf1551 RESA in P. fufciparum, release their contents into the parasitophorous vacuole after the merozoites have entered erythrocytes (Torii et al. 1989). By analogy, Pf155/RESA might be accumulated in the parasitophorous vacuole during invasion and then move from this location to the erythrocyte membrane by an unknown process (Aikawa et al. 1990). By this route of movement, Pfl55/RESA would not be expected to be
exposed and easily accessible to antibodies shown to inhibit invasion very efficiently (Wahlin et al. 1984; Berzins et al. 1986; Ruangjirachuporn et al. 1988; Perlmann et al. 1989). However, the mechanism of this inhibition involving Pf155/RESA as target antigen is presently unknown. In EMIF, antibodies to PflSS/RESA may be seen to stain the surface of erythrocytes to which a merozoite has attached but before it has entered the cell (Perlmann et al. 1984). Moreover, the antigen is also found in easily demonstrable amounts in the medium of P. fulciparum cultures (Perlmann et al. 1984; Carlsson et al. 1990). Taken together, such findings imply that the pathway of antigen release from dense granules in P. falciparum may be different from that postulated for P. knowlesi (Torii et al. 1989),
P.
fakiparum:
ANALYSIS OF ANTIGEN
or, alternatively, that PflWRESA is also present at other locations, not as yet defined by electron microscopy. Further work is needed to resolve these questions. ACKNOWLEDGMENTS This work was supported by grants from the UNDP/ World Bank/WHO Special Programme for Research and Training in Tropical Diseases, the Swedish Medical Research Council, and the Rockefeller Foundation Great Neglected Diseases Network. The support from Kabi AB and SmithKline Beecham Biologicals is gratefully acknowledged.
REFERENCES AIKAWA, M. 1986. Localization of protective malarial antigens by immunoelectron microscopy. Memorius do Institute Oswald0 Cruz 81 Suppl. II, 159-166. AIKAWA, M., TORII, M., SJ~LANDER, A., BERZINS, K., PERLMANN, P., AND MILLER, L. H. 1990. PflSS/RESA antigen is localized in dense granules of Plasmodium falciparum merozoites. Experimental Parasitology 71, 326-329. ANDERS, R. F., MURRAY, L. J., THOMAS, L. M., DAVERN, K. M., BROWN, G. V., AND KEMP, D. J. 1987. Structure and function of candidate vaccine antigens in Plasmodium falciparum. Biochemical Society Symposia 53, 103-l 14. BATTEIGER, B., NEWHALL, V. W. J., AND JONES, R. B. 1982. The use of Tween 20 as a blocking agent in the immunological detection of proteins transferred to nitrocellulose membranes. Journal oflmmunological Methods 55, 297-307. BENNETT, V. 1983. Proteins involved in membranecytoskeleton association in human erythrocytes: Spectrin, ankyrin, and band 3. In “Methods in Enzymology” (S. Fleischer and B. Fleischer, Eds.), Vol. 96, pp. 313-324. Academic Press, San Diego. BERZINS, K., PERLMANN, H., WAHLIN, B., CARLSSON, J., WAHLGREN, M., UDOMSANGPETCH, R., BJ~RKMAN, A., PATARROYO, M. E., AND PERLMANN, P. 1986. Rabbit and human antibodies to a repeated amino acid sequence of a Plasmodium falciparum antigen, Pfl55, react with the native protein and inhibit merozoite invasion. Proceedings of the National Academy of Sciences USA 88, 1065-1069. BERZINS, K., WAHLGREN, M., AND PERLMANN, P. 1983. Studies on the specificity of anti-erythrocyte antibodies in the serum of patients with malaria. Clinical and Experimental Immunology 54,313-318. BIANCO, A. E., CULVENOR, J. G., COPPEL, R. L., CREWTHER, P. E., MCINTYRE, P., FAVALORO, J. M., BROWN, G. V., KEMP, D. J., AND ANDERS, R. F. 1987. Putative glycophorin-binding protein is
PflWRESA
71
secreted from schizonts of Plasmodium falciparum. Molecular and Biochemical Parasitology 23, 91102. BJ~RKMAN, A., HEDMAN, P., BROHULT, J., WILLcox, M., DIAMANT, I., PEHRSSON, P. O., ROMBO, L., AND BENGTSSON, E. 1985. Different malaria control activities in an area of Liberia. Effects on malariometric parameters. Annals of Tropical Medicine and Parasitology 79, 239-246. BROWN, G. V., CULVENOR, J. G., CREWTHER, P. E., BIANCO, A. E., COPPEL, R. L., SAINT, R. B., STAHL, H.-D., KEMP, D. J., AND ANDERS, R. F. 1985. Localization of the ring-infected erythrocyte surface antigen (RESA) of Plasmodium falciparum in merozoites and ring-infected erythrocytes. Journal of Experimental Medicine 162, 774-779. CAPPAI, R., VAN SCHRAVENDIJK, M.-R., ANDERS, R. F., PETERSON, M. G., THOMAS, L. M., CowMAN, A. F., AND KEMP, D. J. 1989. Expression of the RESA gene in Plasmodium falciparum isolate FCR3 is prevented by a subtelomeric deletion. Molecular and Cellular Biology 9, 3584-3587. CARLSSON, J., BERZINS, K., PERLMANN, H., AND PERLMANN, P. 1991. Studies on Pfl55ZRESA and other soluble antigens from in vitro cultured Plasmodium flaciparum. Parasitology Research 77, 2732. CARLSSON, J., UDOMSANGPETCH, R., WI\HLIN, B., AHLBORG, N., BERZINS, K., AND PERLMANN, P. 1990. Plasmodium falciparum: Differential parasite reactivity of antibodies to repeated sequences in the antigen Pfl5YRESA. Experimental Parasitology 71, 314-325. COHEN, R. S., BLOMBERG, F., BERZINS, K., AND SIEKEVITZ, P. 1977. The structure of postsynaptic densities isolated from dog cerebral cortex. I. Overall morphology and protein composition. Journal of Cell Biology 74, 181-203. D~BELI, H., TREZCIAK, A., GILLESSEN, D., MATILE, H., SRIVASTAVA, I. K., PERRIN, L. H., JAKOB, P. E., AND CERTA, U. 1990. Expression, puritication, biochemical characterization, and inhibition of recombinant Plasmodium falciparum aldolase. Molecular and Biochemical Parasitology 41, 259-268. FAIRBANKS, G., STECK, T. L., AND WALLACH, D. F. H. 1971. Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10, 2606-2617. FAVALORO, J. M., COPPEL, R. L., CORCORAN, L. M., FOOTE, S. J., BROWN, G. V., ANDERS, R. F., AND KEMP, D. J. 1986. Structure of the RESA gene of Plasmodium falciparum. Nucleic Acids Research
14, 826.5-8278. FOLEY, M., MURRAY, L. J., AND ANDERS, R. F. 1990. The ring-infected erythrocyte surface antigen protein of Plasmodium falcipnrum is phosphorylated
72
RUANGJIRACHUPORN
upon association with the host cell membrane. MOlecular and Biochemical HOLMQUIST, WIGZELL,
Parasitology
G., UDOMSANGPETCH, H., AND PERLMANN,
38, 69-76.
R., BERZINS, K., P. 1988. Plasmo-
dium chabaudi antigen PchlOS, Plasmodium falciparum antigen Pfl55, and erythrocyte band 3 share cross-reactive epitopes. Infection and Immunity 56,
1545-1550. LORAND, L., BARNES, HAWKINS, M., AND
N., BRUNER-LORAND, J. A., MICHALSKA, M. 1987. Inhibi-
tion of protein cross-linking in Ca*+-enriched human erythrocytes and activated platelets. Biochemistry 26, 308-313. Low, P. S. 1986. Structure and function of the cytoplasmic domain of band 3: Center of erythrocyte membrane-peripheral protein interactions. Biochimica et Biophysics Acta 864, 145-167. NEVILLE, D. M. 1971. Molecular weight determinations of proteindodecyl sulfate complexes by gel electrophoresis in a discontinuous buffer system. Journal
of Biological
Chemistry
246, 6328-6334.
PERLMANN, H., BERZINS, K., WAHLGREN, M., CARLSSON, J., BJ~RKMAN, A., PATARROYO, M. E., AND PERLMANN, P. 1984. Antibodies in malarial
sera to parasite antigens in the membrane of erythrocytes infected with early asexual stages of Plasmodium falciparum. Journal cine 159, 1686-1704.
of Experimenal
Medi-
antigenic diversity in Pfl55, a major parasite antigen in membranes of erythrocytes infected with PlasmoJournal
of Clinical
protein into the plasma membrane of host erythrocytes. Journal of Cell Biology 106, 1507-1513. SCHENKEIN, I., LEVY, M., AND UHR, J. W. 1972. The use of glucose oxidase as a generator of H,O, in the enzymic radioiodination of components of cell surfaces. Cellular Immunology 5, 49&493. SHEETZ, M. P. 1979. Integral membrane protein interaction with Triton cytoskeletons of erythrocytes. Biochimica et Biophysics Acta 557, 122-134. STANLEY, H. A., LANGRETH, S. G., REESE, R. T., AND
TRACER,
Microbiology
25, 2347-2354. PERLMANN, H., PERLMANN, P., BERZINS, K., WAHLIN, B., TROYE-BLOMBERG, M., HAGSTEDT, M., ANDERSSON, I., HUGH, B., PETERSEN, E., AND BJ~RKMAN, A. 1989. Dissection of the human anti-
body response to the malaria antigen Pfl55/RESA into epitope specific components. Immunological Reviews, 112, 115-132. RUANGJIRACHUPORN, W., WAHLIN, B., PERLMANN, H., CARLSSON, J., BERZINS, K., WAHLGREN, M., UDOMSANGPETCH, R., WIGZELL, H., AND PERLMANN, P. 1988. Monoclonal antibodies to a syn-
thetic peptide corresponding to a repeated sequence in the Plasmodium falciparum antigen Pfl55. Molecular and Biochemical Parasitology 29, 19-28. SAM-YELLOWE, T. Y., SHIO, H., AND PERKINS, M. 1988. Secretion of Plasmodium falciparum rhoptry
W.
1982.
Plasmodium
falciparum
merozoites: Isolation by density gradient centrifugation using Percoll and antigenic analysis. Journal of Parasitology 68, 1059-1067. STECK, T. L. 1974. The organization of proteins in the human red blood cell membrane. A review. Journal of Cell Biology 62, l-19. STECK, T. L., WEINSTEIN, R. S., STRAUS, J. H., AND WALLACH, D. F. H. 1970. Inside out red cell mem-
brane vesicles: Preparation and purification. Science 168, 255-257. TAYLOR, D. W., PARRA, M., AND STEARNS, M. E. 1987. Plasmodium falciparum: Fine structural changes in the cytoskeletons of infected erythrocytes. Experimental Parasitology 64, 178187. J. H., MILLER, L. H., AND TORII, M., ADAMS, AIKAWA, M. 1989.Release of merozoite dense granules during erythrocyte invasion by Plasmodium knowlesi.
PERLMANN, H., BERZINS, K., WAHLIN, B., UDOMSANGPETCH, R., RUANGJIRACHUPORN, W., WAHLGREN, M., AND PERLMANN, P. 1987. Absence of
dium falciparum.
ET AL.
Infection
and Immunity
51, 3238-3233.
J. B. 1976. Human malaria parasites in continuous culture. Science 193, 673675. UDOMSANGPETCH, R., LUNDGREN, K., BERZINS, K., PERLMANN, H., TROYE-BLOMBERG, M., CARLSSON,
TRAGER,
W., AND JENSEN,
J., WAHLGREN, M., PERLMANN, P., AND BJ~RKMAN, A. 1986. Human monoclonal antibodies to
Pfl55, a major antigen of malaria parasite PlasmoScience 231, 57-59. UNI, S., MASUDA, A., STEWART, M. J., IGARASHI, I., NUSSENZWEIG, R., AND AIKAWA, M. 1987. Ultrastructural localization of the 150/130kD antigens in sexual and asexual blood stages of Plasmodium falciparum-infected human erythrocytes. American Journal of Tropical Medicine and Hygiene 36, 481488. WAHLIN, B., WAHLGREN, M., PERLMANN, H., dium falciparum.
BERZINS, K., BJORKMAN, A., PATARROYO, M. E., AND PERLMANN, P. 1984. Human antibodies to a M,
155,000Plasmodium falciparum antigen efficiently inhibit merozoite invasion. Proceedings of the National Academy of Sciences USA 81, 7912-7916. Received 6 November 1990; accepted with revision 19 February 1991
