Cell, Vol. 63, 141-153,
October
5, 1990, Copyright
0 1990 by Cell Press
The Duffy Receptor Family of Plasmodium knowlesi Is Located within the Micronemes of Invasive Malaria Merozoites John H. Adams: Diana E. Hudson: Motomi Torii,t Gary E. Ward,’ Thomas E. Wellems, Masamichi Aikawa,t and Louis H. Miller* l Laboratory of Parasitic Diseases National Institute of Allergy and Infectious Diseases Bethesda, Maryland 20892 tlnstitute of Pathology Case Western Reserve University Cleveland, Ohio 44106
Summary Plasmodium vivax and Plasmodium knowlesl merozoites invade human erythrocytes that express Duffy blood group surface determinants. A soluble parasite protein of 135 kd binds specifically to a human Duffy antigen. Using antisera affinity purified on the 135 kd protein, we cloned a gene that encodes a member of a P. knowlesl family of erythmcyte binding proteins. The gene is a member of a family that includes three homologous genes located on separate chromosomes. Two genes are expressed as major membrane-bound products that give rise to soluble erythrocyte binding proteins: the 135 kd Duffy binding protein and a 138 kd protein that binds only rhesus erythrocytes. These different erythrocyte binding specificities may result from sequence divergence of the homologous genes. The Duffy receptor family is localized in micronemes, an organelle found in all organisms of the phylum Apicomplexa.
Duffy-positive blood group determinants on human erythrocytes are essential for invasion of human erythrocytes by Plasmodium vivax, a human malaria, and by Plasmodium knowlesi, a malaria of old-world monkeys that invades human erythrocytes (Miller et al., 1975, 1976). During invasion, a merozoite first attaches to an erythrocyte on any surface of the merozoite, then reorients such that its apical end is in contact with the erythrocyte (Dvorak et al., 1975). Both attachment and reorientation oft? knowlesi merozoites occur equally well on Duffy-positive and -negative erythrocytes (Miller et al., 1975). A junction then forms between the, apical end of the merozoite and the Duffypositive erythrocyte, followed by vacuole formation and entry of the merozoite into the vacuole (Aikawa et al., 1978). Junction formation and merozoite entry into the erythrocyte do not occur on Duffy-negative cells (Miller et al., 1979) suggesting that the Duffy determinant is necessary for apical junction formation but not initial attachment. Soluble proteins that bind to Duffy-positive and not to Duffy-negative human erythrocytes are found in F?knowlesi and P vivax culture supernatants. The P knowlesi Duffy binding proteins also bind to rhesus erythrocytes, but the I? vivax Duffy binding proteins do not. This binding
correlates with the susceptibility of these erythrocytes to invasion by P knowlesi and P vivax. The binding specificities probably reflect the differences in the Duffy blood group antigens of the host to which the parasite has adapted. Using antibody affinity purified with the Duffy binding protein, we have cloned a member of the Duffy receptor gene family of f? knowlesi: Antibodies to two nonoverlapping fusion proteins from the cloned gene immunoprecipitated the F? knowlesi Duffy binding proteins. These proteins were immunolocalized within the micronemes but not on the surface of invasive merozoites. Micronemes are organelles of previously unknown function found within the apical end of all invasive stages of species in the phylum Apicomplexa. Results Cloning of a Member of the Duffy Receptor Gene Family The soluble 135 kd P knowlesi protein that binds specifically to the Duffy blood group antigen was affinity purified with Duffy-positive (Fy a-b+) human erythrocytes, electrophoresed, and electroblotted onto nitrocellulose. Strips of nitrocellulose from the 135 kd region of the gel were used to affinity purify anti-135 kd antibody from immune serum of a rhesus monkey. A Xgtll expression library derived from l? knowlesi asexual erythrocyte-stage cDNA was screened with the anti-135 kd specific antibodies. Two cDNA clones, 2Cl with a 1.2 kb insert and 1Cl with a 2.6 kb insert, were identified and subcloned into plasmids (p2Cl and plC1, respectively). Clone plC1 has 2.2 kb of open reading frame followed by 0.4 kb of untranslated region ending with polyadenylation at the B’end. Two transcripts of 3.8 and 4.2 kb were identified by Northern blot analysis (Figure 1). Since plC1 is only 2.6 kb and contains no start codon at its 5’ end, it is apparent that we have cloned only part of the gene. The sequence of p2Cl is identical to an internal region of plC1 with the exception of one base. To determine if both sequences were present in genomic DNA, two 17 bp oligonucleotides that had the base from p2Cl or plC1 in position 9 were used to probe restriction digests of f? knowlesi genomic DNA. The oligonucleotide probe from plC1 hybridized to the three fragments in an EcoRl digest that were hybridized by the cDNA clones (see below). The probe from p2Cl did not hybridize at the same stringency, indicating the one base pair difference in p2Cl was a cloning artifact or transcriptional error. An oligonucleotide (oligo 13; Figure 2) from the 5’ end of the cDNA clone plC1 hybridized only to a 6 kb EcoRl genomic fragment (Figure 4C) and was used to clone this fragment (pEco6). From the 5’ end pEco6 had sequence of 113 bp not present in plC1 followed by sequence that was identical to plC1 for the next 150 bp. Additional sequence unique to plC1 and the 6 kb EcoRl genomic fragment was identified with oligonucleotide probes 46 and 52
(Figure 5). At the 3’ end of the open reading frame of pEco6, there were three introns identified as defined by genomic sequences that were not present in the cDNA sequences (Figure 2). The 5’ and 3’ borders of the introns (GTA. YAG) were identical to consensus splice sites for other malaria and eukaryotic genes (Weber, 1966; Darnell et al., 1966). Comparison of internal restriction fragments (Haelll and Ndel) indicated no additional introns or size differences between plC1 and pEco6 (data not shown). The deduced amino acid sequence of the C-terminal portion of the gene, which covered four exons (Figure 2), shows that the gene has a 22 amino acid transmembrane segment followed by 45 amino acids at the C-terminus. The presence of a transmembrane domain is consistent with the function of a receptor molecule. Nine repeats of the pentapeptide SSD(Q/H)T occur 5’ to the transmembrane segment. Two regions of high cysteine content are separated by a proline-rich region. There is no significant sequence identity of either the genomic or cDNA clones with any gene or protein in EMBL 21 or Swiss-Prot 13 data bases, respectively (Intelligenetics). The cDNA clones p2Cl and plC1 hybridized with three chromosomes of 3.6, 1.6, and 1.2 Mb separated in a pulsed-field gel electrophoresis (Figure 3 and data not shown), indicating that three cross-hybridizing elements in the l? knowlesi genome were recognized by the cDNA probe. Two subregions of plC1 (DR.1 and DR.2; Figure 2) were used to probe restriction digests of genomic DNA. The probes did not cross-hybridize with each other (data not shown). Probe DR.1 hybridized with three bands in EcoRl, Nsil, and Ndel digests (Figure 4A). Probe DR.2 hybridized with bands of identical mobility to those hybridized with DR.1 in the EcoRI, Nsil, and Ndel digests plus three additional bands in the Ndel digest, including an 600 bp band known to be found in the DR.2 sequence (Figure 48). Hybridization with at least three bands in these digests is consistent with the hybridization to three chromosomes observed with the pulsed-field gel electrophoresis and suggests homology in the three gene fragments in both the 5’ (DR.2) and 3’ (DR.l) regions of plC1. Southern blot analysis of the three cross-hybridizing regions of R knowlesi and P vivax genomic DNA was performed with oligonucleotide probes from the sequence of pEco6 (Figure 5). The probes hybridized with all three of the P. knowlesi EcoRl fragments at low stringency (see Experimental Procedures). At higher hybridization stringencies, some of the oligonucleotide probes hybridized to only one EcoRl restriction fragment (6 kb), two EcoRl restriction fragments (6 and 10 kb or 6 and 4 kb), or equally to all three EcoRl restriction fragments (Figure 5). The failure of an oligonucleotide to hybridize to an EcoRl fragment was not due to the position of the EcoRl site because the oligonucleotide also did not hybridize with the corresponding fragments using other restriction enzymes. Furthermore, they did hybridize with all three EcoRl fragments at lower stringencies. The analysis with oligonucleotide probes revealed that the 5’ends of the homologous P. knowlesi genes are divergent. In the 5’ portion of pEco6 only one of five oligonucleotide probes hybridized at high stringency with the 4 kb
9.57.54.43
0 F X 4 Y
2.41.35-
0.78-
Figure 1. Identification Family
of RNA Transcriptsof
the Duffy Receptor
Gene
Poly(A)-enriched RNA from late-stage schizonts of P knowlesi was separated by agarose gel electrophoresis (1% agarose, 20 mM MOPS, 5 mM sodium acetate, 0.5 mM EDlA, 200 mM formalin), transferred onto GeneScreen Plus in 20x SSC, cross-linked onto the membrane with ultraviolet, and dried under vacuum. Insert of plC1 was radiolabeled by the random priming method and hybridized to two closely migrating transcripts of 3.8 and 4.2 kb in 6x SSC, 20 mM HP04 (pH 6.6) 5x Denhardt’s solution, 0.5% SDS, 100 ug/ml sodium heparin, and 50 uglml sheared salmon sperm DNA at 65OC overnight, and had a final wash in 0.1x SSC, 0.1% SDS at 55%.
EcoRl fragment and three of five hybridized with the 10 kb EcoRl fragment, but four of four hybridized with P vivax genomic DNA (Figure 5). These data are consistent with the fact that the 6 kb EcoRl fragment is most similar to the equivalent single-copy Duffy receptor gene present in the P vivax genome (Fang et al., 1990). The oligonucleotides from the central region (50, 52, and 54) were not hybridized to P vivax because the sequence of the P vivax Duffy receptor is nonhomologous in this region (Fang et al., 1990). In the 3’ region of the gene (oligonucleotides 56 to 35; Figure 5), there was a high degree of homology among the three P. knowlesi genes. To determine whether the different RNA species of l? knowlesi were products of alternative splicing of a single gene or transcripts of different genes, we used various oligonucleotides to probe Northern blots. Oligonucleotide probes from the three introns did not hybridize to Northern blots but drd. hybridize to the 6 kb EcoRl fragment on Southern blots (data not shown). An oligonucleotide probe
~~;modium
knowlesi
A
Duffy
Receptor
Family
B Via
Vlc
Via
overlapping the possible splice site between the fourth exon also did not hybridize to a Northern knowlesi poly(A) RNA (data not shown). The data that the two transcripts originated from two genes from alternative splicing of a single gene.
Vlc
+
lmmunochemical Studies of the Erythrocyte Binding Proteins We determined if antisera raised to portions of the cloned gene immunoprecipitated erythrocyte binding proteins of F? knowlesi. These soluble proteins are divided into two groups: proteins of 135 kd and 120 kd that bind to human Duffy-positive erythrocytes and rhesus erythrocytes, and those that do not bind human erythrocytes but bind to rhesus erythrocytes. Antisera to fusion proteins of nonoverlapping regions of
-b + :.,i::i..,, :. Figure somes
3. Genes
of the Duffy
Receptor
first and blot of l? indicate and not
Family
Are on Three
Chromo-
(A) Chromosomes prepared from two f? knowlesi clones Vla and Vlc were separated by pulsed-field gel electrophoresis and stained with ethidium bromide. (6) Insert of p2Cl was radiolabeled by the random priming method and hybridized chromosomes of three sizes (1.2 x 106, 1.6 x 106. 3.6 x
10s bp) in 1 M NaCI, 1% SDS, 0.05% nonfat dry milk, 56 mM Tris (pH 6.0) and 100 kg/ml sheared salmon sperm DNA at 65% overnight and a final wash in 0.2x SSC, 0.1% SDS at 55oC. Chromosome sizes were determined previously relative to their migration to Plasmodium falciparum chromosomes (Hudson et al., 1966).
Cell 144
A
B
Figure 4. Identification Gene Family
of Restriction
Fragments
of the Duffy
-10
g
-6 -4
; ”
Receptor
Restriction digests (Nsil, Bglll, Ndel, and EcoRI) of I? knowlesi genomic DNA were separated by agarose gel electrophoresis and blotted onto GeneScreen Plus (Du Pont). Nonoverlapping crosshybridizing fragmentsof plC1, DR.1, and DR.2 (see Figure 2) were amplified by the PCR and radiolabeled by the random priming method. DR.1 (A) and DR.2 (B) were used to probe the same Southern blot of the P knowlesi genomic DNA digests using hybridization conditions as described in Figure 2. Oligonucleotide 13 (C) (see Figure 2) hybridized to a single genomic fragment with each of the restriction digests examined. The oligonucleotide, radiolabeled by the 5’ terminus labeling method, was hybridized in 6x SSC, 20 mM HP04 (pH 6.6) 5x Denhardt’s solution, 05% SDS, and 100 ug/ml sodium heparin at 42OC and then washed in 6x SSC, 0.5% SDS repeatedly at room temperature and then for 20 min at 46OC. Blots were stripped with 0.5 M NaOH, 1.5 M NaCl between hybridizations. Molecular sizes were calculated from known restriction digest fragments of &DNA (Hindlll) and cpX 174 RF DNA (Haelll).
the Duffy receptor (DR.1 and DR.2) immunoprecipitated the 135 kd Duffy binding protein from culture supernatants. The immunoprecipitated 135 kd protein was removed by preadsorption with Duffy b human erythrocytes,
but not by Duffy-negative erythrocytes (Figures 6 and 7; Table 1). The immunoprecipitated 135 kd protein was also adsorbed by erythrocytes of old-world (rhesus) and newworld (Aotus and Cebus) monkeys, which are Duffy positive. The 120 kd Duffy binding protein was precipitated by anti-DR.1 (Figures 6 and 7) and by anti-peptide 3 serum but not by anti-DR.2 serum (data not shown). The fact that antisera to different regions of plC1 immunoprecipitated the 135 kd and the 120 kd Duffy binding proteins indicates that the clone encodes the Duffy receptor or a member of the Duffy receptor family. The DR.1 and DR.2 antisera also immunoprecipitated soluble proteins from l? knowlesi culture supernatants that bound to rhesus erythrocytes but not to human erythrocytes. A 136 kd protein, which was adsorbed only by rhesus erythrocytes, was immunoprecipitated by both anti-DR.1 and anti-DR.2 sera (Figure 7). Proteins of 125 and 160 kd, which were also specifically adsorbed by rhesus erythrocytes, were only immunoprecipitated by antiDR.1 serum (Figures 6 and 7). A 155 kd protein that was adsorbed by Aotus erythrocytes and partially adsorbed by rhesus erythrocytes (Miller et al., 1966) was immunoprecipitated by both anti-DR.1 and anti-DR.2 sera. In some preparations a 153 kd protein that was not adsorbed by any erythrocyte in our assay was immunoprecipitated by both DR.1 and DR.2 antisera. The antiserum to the C-terminal peptide did not immunoprecipitate any of the soluble erythrocyte binding proteins, but it did immunoprecipitate the membrane-bound proteins and the proteins from in vitro translated mRNA that were seen by the anti-fusion protein serum (see below). As the C-terminus is 45 amino acids from the putative transmembrane domain, the failure of this antiserum to immunoprecipitate the soluble proteins is consistent with the proteolytic cleavage site for the soluble proteins being located amino to the transmembrane domain. Pulse-Chase Analysis of Membrane-Bound Proteins To identify membrane-bound proteins that may be precursors to the soluble proteins, we immunoprecipitated detergent-solubilized parasite proteins. From Briton X-100 extracted parasites, anti-DR.l, anti-DR.2, and anti-c-terminus sera immunoprecipitated a closely migrating doublet of 146 and 145 kd and a minor protein of 170 kd (Figure 3
5 2913 4646 ~ligonuckotide
49
50 52 54
56
57 56 16 105 106
1075935
probe
Figure 5. Structure of the 6 kb EcoRl Genomic Fragment and Sequence Relatedness to Other Members of the P. knowlesi Gene Family
The exons of the predicted open reading frame of the 6 kb EcoRl fragment are shown as boxes and the introns as lines. The B’end is shown as a jagged line to indicate incomplete sequence ++++ + + ++ + + +++ + +++ in this region. Oligonucleotide probes derived from the 6 kb EcoRl P knowlesi gene fragment -mm+were used to probe genomic digests of f? + - + ++ +-+ + +++ knowlesi and I? vivax Southern blots. Genomic DNA was digested with multiple restriction enP. vivax ND+++ + ND ND ND ND ND NDND + ND NDND+ zymes (P knowlesi: EcoRI, EcoRI-BamHI, EcoRI-Dral, Nsil, Nsil-BamHI, Nsil-Dral; P vivax: EcoRI, Hindlll, Dral, Kpnl), separated by agarose gel electrophoresis, denatured, and blotted onto GeneScreen Plus (Du Pont). Oligonucleotides were radiolabeled and hybridized as described in Figure 3. Melting point (T,,,) was estimated using the formula T, = (% GC) (0.41) + 61.5 (675/[number of bases in oligonucleotide probe]). Blots were stripped with 0.5 M NaOH, 1.5 M NaCl between hybridizations. +--++
-
-
-
++
+++
+
- ++
yla;modium
knowlesi
Duffy
Receptor
Family
A
m E
$ul
d 2
lmmunoprecipitated with anti-DR.1
m a 2
B EBA
k
EBA
-
lmmunoprecipitated with anti-DR.1
sz
zz
Preadsorbed
C
EBA Jz a
IL
D ha
AZ
j
specific for the following reasons. First, the sense oligonucleotide did not block synthesis of the three proteins (Figure 9B). Second, the higher molecular weight bands seen in immunoprecipitates with anti-DR.2 (Figure 9A) were unaffected by the antisense oligonucleotide (data not shown). Third, the total incorporated trichloroacetic acidprecipitable [%S]methionine counts were the same for samples with and without the antisense oligonucleotide (data not shown). The data indicate that a family of genes is expressed that yields a family of erythrocyte binding proteins, at least one of which is a Duffy binding protein. We refer to this family of erythrocyte binding proteins and their membranebound precursors as the Duffy receptor family.
lmmunoprecipitsted with anti-DR. 1
- 200
-
1138 -135 -125 -120
-
Figure 7. Analysis of l? knowlesi Soluble Rhesus-Specific Binding Proteins of the Duffy Receptor Family
Localization of the Duffy Receptor Family in Merozoites Antisera to the Duffy receptor family proteins permitted the study of their localization in fixed, detergent-permeabilized schizont-infected erythrocytes and merozoites (Figure 10). These proteins were first detectable late in schizont development; a diffuse fluorescence developed at the apical end of developing merozoites at the 8+ nuclei stage of schizonts (Figure 1OB). Strong discrete fluorescence appeared at the apical end when the merozoites were fully formed. At this stage the hemozoin pigment had coalesced into a single refractile granule, an event that occurs just before rupture. A strong spot of fluorescence was also seen at the apical end of free merozoites. Control antisera showed only faint background staining of infected ceils (Figure 10D). Only fixed and permeabilized parasites showed positive immunofluorescence; the Duffy receptor family proteins were not found on the surface of intact, invasive merozoites. The precise location of the Duffy receptor family was determined by immunoelectron microscopy. The Duffy receptor family is localized in the micronemes of late schizonts and free merozoites (Figure 11). This same localization was found with anti-DR.1, anti-DR.2, and anti-C-terminus sera and was independently confirmed by the ferritinbridge technique using anti-DR.1 and anti-DR.2 sera (Figure 11D and data not shown). No detectable immunolabeling of merozoites was seen using control antisera (data not shown).
Erythrocyte
Metabolically labeled l? knowlesi culture supernatant was untreated or preadsorbed with 2 vol of packed cell volumes of washed erythrocytes of human Duffy-negative (Fy Neg), human Duffy b-positive (Fyb), and rhesus erythrocytes (Rh), instead of 1 vol of packed cells as in Figure 6, and immunoprecipitated with anti-DR.1 rabbit serum. Erythrocyte binding proteins of rhesus, human Duffy b, and human Duffy-negative erythrocytes were affinity purified in erythrocyte binding assays (EBA) from the same culture supernatants as the immunoprecipitated proteins and were electrophoresed on the same SDS-PAGE gel. The relative molecular sizes of the immunoprecipitated proteins were calculated from prestained molecular size standards of 200, 97.4,66, and 43 kd (Bethesda Research Laboratories) and adjusted relative to the 135 kd protein. The SDS-PAGE gel contained 0.6% bis cross-linker and mixed-length SDS and was 0.75 mm thick.
Table
1. Soluble
Soluble
Protein
Erythrocyte
Binding
Erythrocyte
120 125 135 136 155 160
(kd)
Human
Proteins
Binding
Duffy+
+ -
Culture
Supernatants Precipitating
Antiserab
Rhesus
Aotus
Anti-DR.1
Anti-DR.2
+ + + + + +
+
+ + + + + +
Specificitya
Human
+
from P. knowlesi
Duffy
a Plus signs = adsorbs; minus signs = does not adsorb. b Plus signs = immunoprecipitated; minus signs = not immunoprecipitated.
+ +
The locations
Anti-C-Terminus
Anti-Peptide
3
+ + + +
-
of DR.1, DR.2, and the peptides
-
are described
in Figure
1.
~la;modium
knowlesi
Duffy
Receptor
Family
A
.
B
u-DR.1 921
t :3 F? 0x 97.4-
1Wk 170-+ 166-
5 68-
TX-100
pellet
supernatarlts
Figure 9. In Vttro Trenslatttn of Duffy Receptor Gene and Their Inhibition with Antisense Oligonucleotides
Family Products of p&o8
(A) In vitro translation products of late schizont t? knowlesi RNA were precipitated with anti-DR.1 (a-DR.l), anti-DR.2 (a-DR.2) anti-Cterminus (a-C-tail), anti-peptide 3 (a-Pep3), and anti-wild-type trpE (a-trpE). (B) Late schizont RNA was annealed with oligodeoxynucleotide (sense strand [oligo 30; Figure 21 ATGGGAACTAATATGGA, antisense [oligo 311 TCCATATTAGTTCCCAT) at 3pC for 20 min, treated with RNAase H, translated in vitro, and immunoprecipitated with anti-DR.1. The molecular sizes of the Duffy receptor family precursors (180, 170, and 165 kd at arrows) were calculated from prestained molecular size markers (as marked for each gel) and adjusted relative to the 135 kd protein run on the same gel. The SDS-PAGE gels contained 0.8% bis cross-linker and mixed-length SDS and was 0.75 mm thick. SDS pellet Figure 6. Pulse-Chase Proteins in Merozoites
Analysis
supernatants of P knowlesi
Duffy
Receptor
Family
l? knowlesi cultures were metabolically labeled with [ssS]methionine/cysteine for 60 min, washed, and then chased in RPM 1640 without additional metabolic label for 0, 90, 160, 270, and 360 min (A and 6) or 0,30,64 90, and 540 min (C and D) (minutes of chase are shown over each lane). Pellets and supernatants were separated at each time point by centrifugation. All immunoprecipitations shown were done with anti-DR.1. The major merozoite-associated proteins of 145 and 148 kd are identified, and a minor protein of 170 kd is indicated by an arrowhead (A and C). (A) Pellets were extracted in Triton X-100 (1% Triton X-100 in 10 mM HEPES [pH 6.51, 50 uglml chymostatin, 50 us/ml leupeptin). (6) Supernatants from the cultures used for the Triton X-100 pellets. (C) Pellets were extracted by boiling in SDS (0.5% SDS, 50 mM Tris [pH 7.41, 100 mM NaCI, 2 mM EDTA), cooled, and mixed with protease inhibitors, DNAse I and Triton X-100 (to 2%). (D) Supernatants from the cultures used for the SDS-extracted pellets. These supernatants were first mixed with SDS (to 05%) and then Briton X-100 (to 2%) before immunoprecipitation. Erythrocyte binding assays (EBA) were done using supernatants from 9 hr cultures and were run in the right-hand lanes of(B) and (D). The relative molecular sizes of the erythrocyte binding proteins and immunoprecipitated proteins were calculated from prestained molecular size standards of 200,97.4,66, and 43 kd (Bethesda Research Laboratories) and were adjusted relative to the 135 kd protein. All of these SDS-PAGE gels contained 0.6% bis cross-linker and mixed-length SDS and were 1.5 mm thick.
Discussion The human Duffyblood group antigen is required for invasion of human erythrocytes by f? knowlesi and l? vivax. Soluble proteins of 135 kd and 120 kd released into the supernatant of P knowlesi cultures bind specifically to Duffy-positive erythrocytes. Using antibody to the 135 kd protein, we have cloned a gene encoding a member of a family of highly homologous proteins that are antigenitally cross-reactive. Antisera to fusion proteins and a peptide derived from sequences within the gene immunoprecipitated two groups of erythrocyte binding proteins: the 135 kd and 120 kd Duffy binding proteins that bind to human and rhesus erythrocytes, and proteins between 160 and 125 kd that bind to rhesus but not to human erythrocytes. The selective binding of some soluble protein8 of t? knowlesi to rhesus erythrocytes is expected, as the rhesus is the host to which the parasite is adapted. Do these proteins bind to the Duffy determinant on rhesus erythrocytes? Because there is no known Duffy-negative rhesus erythrocyte as there is in humans (Marsh, 1975; Mason et
Cdl 143
Figure
10. lmmunofluorescent
Localization
of the Duffy
Receptor
Family
in Schizonts
Blood film smears of mature schizonts were fixed with formalin, permeabilized with 0.1% Triton X-100, and incubated with anti-DR.1 (A and 6) or control serum (C and D) (anti-wild-type trpE), followed by rhodamine-conjugated goat anti-rabbit IgG. The immunoffuorescence pattern is shown in (B) and (D), and the corresponding phase-contrast images in (A) and (C). Free merozoites (1) and fully formed merozoites within mature schizonts (2) show a discrete spot of fluorescence at their apical end. Less mature schizonts (3) show diffuse apical fluorescence, and early schizonts (4) show no staining above background (D). The anterior and posterior ends of each merozoite were determined by staining the nucleus with a DNAspecific fluorochrome, bisbenzimide If33258 (data not shown). A pattern of fluorescence similar to that shown here was seen using anti-DR.2 and anti-C-terminus sera (data not shown)
al., 1977; Hadley et al., 1986) we are limited to indirect data from rhesus erythrocytes in which the Duffy blood group antigen was enzymatically destroyed by chymotrypsin treatment (Miller et al., 1975). None of the soluble proteins bound to these Duffy-negative rhesus erythrocytes (Haynes et al., 1988). If these proteins do bind to Duffy blood group determinants on rhesus erythrocytes, why then do they not bind to human Duffy-positive erythrocytes? We speculate that the Duffy binding protein of f? knowlesi binds better to the rhesus Duffy determinants. For example, the Duffy binding proteins of F?vivax bound to human Duffy-positive and not to human Duffy-negative erythrocytes, but they did not bind to rhesus erythrocytes (Wertheimer and Barnwell, 1989). The likely explanation of these data is that the Duffy blood group determinants on rhesus and human erythrocytes are different and that the parasite receptors recognize these differences. Although erythrocytes from rhesus are Duffy b, that is, reactive with anti-human Duffy b, they do not react with anti-Fy6, a monoclonal antibody that binds to all human Duffy-positive erythrocytes (Nichols et al., 1987). The fact that antisera to the two fusion proteins immunoprecipi-
tated rhesus erythrocyte binding proteins, at least two of which are Duffy binding proteins, indicates that we have cloned the gene encoding the Duffy receptor or a highly homologous gene from a gene family. Expression of a gene family is suggested by the finding of three homologous regions in the f? knowlesi genome. The two transcripts observed on Northern blot hybridization can result from either transcription of multiple genes or alternative splicing of a single transcript. The data from Northern blot analysis and in vitro translation indicate expression from at least two genes. First, the transcripts did not hybridize with antisense oligonucleotides from the three introns. Second, all in vitro translation products that are immunoprecipitated by the anti-fusion protein sera are also immunoprecipitated by anti-C-terminus serum, indicating that alternative splicing cannot delete the C-terminus. The only possible alternative splicing that would maintain the open reading frame of the C-terminus would be between the first and fourth exons. An oligonucleotide overlapping the possible splice site between the first and fourth exon did not hybridize with the Northern blot. Additional evidence for expression of multiple genes
Plasmodium 149
Figure
knowlesi
Duffy
11. lmmunoelectron
Receptor
Microscopic
Family
Localization
of the Duffy
Receptor
Family
(A and 8) Extracellular merozoite fixed in EAI-formaldehyde followed by EGS. Duffy receptors revealed by anti-DR.1 serum and colloidal gold are localized over the micronemes (MN). Rhoptries (R) and dense granules (D) are not immune labeled. N indicates a nucleus. (C) Apical end of the extracellular merozoite fixed in EAI-formaldehyde followed by glutaraldehyde. Duffy receptors revealed by anti-C-terminus serum and colloidal gold are localized over micronemes (MN). (D) Ferritin-bridge immunoelectron microscopy of a merozoite with anti-DR.1 serum used as a primary antibody. Ferritin particles are seen along the edges of micronemes (MN). All bars equal 0.2 urn. No detectable immunolabeling of parasites was seen with control sera (prebleed serum for each antiserum and serum from a rabbit immunized with wild-type trpE protein: data not shown).
comes from in vitro translation and pulse-chase experiments. Three translation products immunoprecipitated by antisera to the fusion proteins and the C-terminus, and the hybrid arrest of these in vitro translation products by antisense DNAfrom the cloned sequence are also indicative of expression of multiple genes. In pulse-chase experiments, a 148/145 kd doublet was immunoprecipitated from detergent-extracted parasites. The relative intensity of the two proteins remained constant during the chase period, indicating that the two were not the result of one protein
being processed to the other. During the chase period, a soluble 138/135 kd doublet appeared in the culture supernatant. These soluble proteins are likely to be derived from the 1481145 kd proteins by cleavage amino to the transmembrane region, since these soluble proteins were no longer precipitated by the anti-C-terminus. The soluble 135 kd protein bound to the Duffy determinant and the 138 kd protein bound to an unidentified rhesus erythrocyte determinant. Soluble erythrocyte binding proteins of greater than 148 kd may have derived from the third gene. Thus,
Cell 150
two or three genes in I? knowlesi express a family of erythrocyte binding proteins that we refer to as the Duffy receptor family. The Duffy receptor gene from F?vivax has been cloned (Fang et al., 1990), and unlike l? knowlesi, the data are consistent with a single-copy gene for the l? vivax Duffy receptor. The i? vivax sequence is highly homologous to the cloned F? knowlesi gene, including the sequence in the cysteine-rich regions, the location and sequence of the three introns, and the amino acid sequence of exons 2, 3, and 4 (Fang et al., 1990). The positions of the cysteines are conserved, which suggests that the tertiary structures of the two proteins are similar. When we mapped the three F?knowlesi genes with oligonucleotides from pEco6, the 5’ probes (oligonucleotides 29 to 49; Figure 5), which include one of the cysteine-rich regions, were more similar to the P vivax gene than to the other P. knowlesi genes. We speculate that this is the erythrocyte binding domain and divergence in this region is the basis for the differing receptor specificities. The antisera that immunoprecipitated the Duffy binding proteins and the related erythrocyte binding proteins were used to immunolocalize the proteins in schizonts and merozoites. These proteins were expressed very late in schizont development and localized in the micronemes. The antiserum to the C-terminal peptide also bound to the micronemes. This anti-peptide serum immunoprecipitated the same products from in vitro translation as the two antisera to the fusion proteins, which indicates that the sequences recognized by these antisera are part of the same protein. Since the antisera were raised to sequences from both sides of the transmembrane region and all bound to micronemes, it appears that the proteins in micronemes are membrane bound. Micronemes are electron-dense membrane-bound organelles found in the apical end of invasive stages of all species of the phylum Apicomplexa (e.g., Toxoplasma, Cryptosporidium, and Babesia). Another protein known to be localized within micronemes is the circumsporozoite protein of malaria sporozoites (Fine et al., 1964). The circumsporozoite protein is believed to be a receptor molecule, as defined by conservation of regions I and II in all species of malaria parasites (Dame et al., 1964). Region II is found in adhesion molecules in humans (e.g., thrombospondin; Kobayashi et al., 1966; Lawler and Hynes, 1966). A peptide derived from region I binds liver cells with high affinity (Aley et al., 1966). The function of other proteins localized in micronemes is unknown (Uni et al., 1967; Cochrane et al., 1969). Thus, two molecules that appear to function as receptors are found in micronemes, suggesting that micronemes may function in the trafficking or sequestration of cell surface receptors. The Duffy receptor family proteins were not found on the surface of viable, invasive merozoites. These proteins may be released from the micronemes to the merozoite surface after initial attachment and apical reorientation. This would be consistent with the observation that attachment and apical reorientation occurs equally well to Duffypositive and -negative human erythrocytes, but junction formation only occurs between merozoites and Duffy-
positive erythrocytes. The sequestration of receptors in micronemes until after the merozoite attached to the erythrocyte would reduce exposure of the receptor to immune sera and may be an additional mechanism of immune evasion. ExperImental Pnxedum Metabolically
Labeled
Parasites
P. knowlesi (Malayan H) schizont-infected erythrocytes (2 x lO’/ml) were metabolically labeled with 75 WCilml [35S]methionine/cysteine (ICN Radiochemicals) in methionine/cysteine-deficient RPM 1640 culture medium (30 mM HEPES. 0.2% dextrose, 5 mgll hypoxanthine, 0.225% NaHC03) containing 2% fetal bovine serum or were metabolically labeled with 50 kCi/ml [35S]methionine (Amersham) in methionine-deficient RPM1 1640 culture medium. The parasites were cultured 9-l 1 hr at 3pC lo allow complete rupture and release of merozoites. Culture supernatants were centrifuged at 20,000 x g for 20 min before freezing at -70°C (Haynes et al.. 1988).
Erythmcyte
Binding
Aseay
The erythrocyte binding assay was performed as described previously (Haynes et al., 1988). Briefly, washed erythrocyteswere incubated with culture supernatants (1 vol of erythrocytes lo 4 vol of culture supernatant), passed through silicone oil (GE Versilube F50), washed quickly in RPM1 1640, and passed through silicone oil again. Molecules adsorbed onto the erythrocytes were eluted in a final concentration of 300 mM NaCl(20 @Iof 1.5 M NaCl to 80 ~1 of packed erythrocytes). For the experiments in Figure 6, the wash steps were omitted to increase the detection of the poorly adsorbed proteins. The eluted material was mixed 1:l with SDS-PAGE sample buffer and electrophoresed.
Library
Scmening
The 135 kd Duffy binding protein was isolated from P. knowlesi culture supernatants by large-scale preparations of the erythrocyte binding assay using Duffy-positive (Fy a- b+) human erythrocyles. Eluded molecules were partially purified by SDS-PAGE and electroblotted onto nitrocellulose. The position of the 135 kd antigen was marked by immunolabeling the edges of each blot. Serum from an immune rhesus monkey (number 626; Miller et al., 1977) was diluted to 1% in 0.05% Tween-20, phosphate-buffered saline (pH 7.4) (TPES), and incubated with nitrocellulose strips of the 135 kd protein. Strips were washed three times in TPBS. Bound antibodies were eluted in 100 mM glycine, 150 mM NaCl buffer (pH 2.8), then neutralized with 2 M Tris (pH 8.0) and dialyzed with TPBS containing 0.05% NaN3 (Torii et al.. 1989). The polyclonal antibody purified with the 135 kd protein by this method reacted predominantly to the 120 and 135 kd Duffy binding proteins as well as to the 125. 138, 155, and 160 kd erythrocyle binding proteins (data not shown). A size-selected (>800 bp) amplified P knowlesi lgtll cDNA expression library was constructed from late schizont mRNA (Hudson et al., 1988). Recovered monospecific polyclonal antibodies to the 135 kd Duffy binding protein and alkaline phosphatase-conjugated goat anti-human IgG (H+L; Promega Biotech) were used to screen the i? knowlesi Igtll cDNA library (Young and Davis, 1983). A genomic library was constructed in pUC13 from EcoRl-digested I? knowlesi (clone A, Malayan H) DNA. The 6 kb EcoRl fragment was cloned by selection with oligonucleotide 13 (GGGGATCCGGGAACTGATGAAAAGGCCAAG; Figure 1). Clones were selected by colony hybridization with oligonucleotide 13 using a final washing stringency of 48OC in 6x SSC. 0.5% SDS for 20 min.
Subcloning
and Southern
Blot Hybridization
Conditions
The lgtll cDNA clones, llC1 and 12C1, were subcloned into plasmid vectors pUC13 and Bluescript KS in Escherichia coli strain DH5a (Bethesda Research Laboratories), respectively. Plasmid was purified by conventional techniques from plasmid-transformed cells. Both strands of each of the clones were sequenced by the dideoxy termination method using synthetic oligonucleotide primers (Synthecell Corporation) and T7 DNA polymerase (US Biochemical Corporation) on denatured double-stranded DNA. P knowlesi genomic DNA was prepared for pulsed-field gel electrophoresis from schizont-infected rhesus erylhrocytes using clones Vla and Vlc (Hudson et al., 1988).
$smodium
knowlesi
Table 2. Oligonucleotides
Duffy
Receptor
Used
Oligo
Position
T&C)
29 13 48 48 49 50 52 54 56 57 58 16 105 108 107 59 35
51-85 117-137 143-169 414-439 574-800 1018-1041 1154-1177 1318-1341 1473-1493 1697-1719 1870-l 893 1949-l 970 2131-2148 2316-2333 2830-2847 2709-2738 2748-2764
79 89 70 64 69 77 72 74 63 65 72 64 87 60 62 73 56
Family
for Southern Estimate
See Figure 5 for description of oligonucleotides. wash temperature used. a Sequence includes polylinker.
Hybridization Final Wash
(“C)
87 57 60 57 59 67 62 64 57 55 62 58 57 50 52 65 45
Sequence AAGTCGACTTGGTGGGAATACCCAATACCTTCCATATTAGCCCAT~ GGGGATCCGGGAACTGATGAAAAGGCCAAGa T-fAGATTCATTCCACCATTGTl-TACGA GTTATCCATTGATCATATGATTTAAC CACGl-f-TGTTAAT-fTCATTCTCAAAAG TACCATCTCCACCCGCAGCACCAT TCCTTAGGTXAGTATCAGTGGCA TGAGAGTACTACCTCCTGCGGTTT CGAAGCTTACCGTGAGTTGTAl-TATCAATa GCAAGCTTCTGATCAAACTATAGATACAG” CCCTTGTTGCTTTGACATCTCTGT GTCTTCTATAGTGTTACAGTAT GCTCCACAGGTATACAGG CTCGAATGAGTAACCAAA TGCCTAACGGTAATTCAA TlTCl-TCTTTAGATATTGAGCACATGCAAC CCCTGGATTATTCATAA
T, estimate
= (% GC) (0.41)
For Southern blot hybridizations, genomic or plasmid DNA was digested with restriction enzymes (1 Wug) and separated by agarose gel electrophoresis and visualized by ethidium bromide staining. The DNA was denatured in 0.5 M NaOH, 1.5 M NaCl followed by 1 M NH&HsOz, 1.5 M NaCl and transferred overnight onto GeneScreen Plus (Du Pont), and the filters were baked for 2 hr at 8ooC in vacuum. Oligonucleotide probes were radiolabeled in a forward reaction T4 polynucleotide kinase with [y-32P]ATP (130 uCi160 ng) and the enzyme was heat inactivated at 95OC before use. DNA fragments (plC1, p2C1, DR.1, DR.2) were radiolabeled by random priming with hexanucleotides and the large fragment of DNA polymerase I (Klenow) (Bethesda Research Laboratories) overnight at 25oC, and the enzyme was inactivated at 95OC before use. Hybridizations with oligonucleotides were done overnight in 6x SSC, 20 mM HP04 (pH 6.8) 5x Denhardt’s solution, 0.5% SDS, and 100 pg/ml sodium heparin at 3pc and then washed in 6x SSC, 0.5% SDS repeatedly at room temperature and then for 20 min at 3pC. Subsequent washes were done by increasing increments of 5°C. Melting point (T,) was estimated using the formula T, = (% GC) (0.41) + 81.5 - (675/[number of bases in oligonucleotide probe]) (Davis et al., 1988). Hybridization with the DNA fragments was done overnight in the same buffer as the oligonucleotide probes or in 1 M NaCI, 1% SDS, 0.05% nonfat dry milk, 50 mM Tris (pH 8.0) and 100 rg/ml sheared salmon sperm DNA at 65OC overnight and washed in 2x SSC at 55V. Final washing conditions for the results shown are given in the figure legends. Table 2 lists the sequence and its position in pEco8, estimated melting point, and final wash temperatures for the oligonucleotide probes described in Fig ure 5. Blots were stripped with 0.5 M NaOH, 1.5 M NaCl between hybridizations. RNA Purlflcation and Analysis The first four preparations of RNA were from late-stage schizonts of P knowlesi isolated by centrifugation in large hematocrit tubes and cultured 3-4 hr in RPM1 1640 culture medium with or without 50 ug/ml chymostatin and 50 ug/ml leupeptin. RNA was extracted using the hot phenol method described previously (Wellems and Howard, 1988). For the fifth and sixth preparations, white blood cells were removed from the parasitized blood using a Sepacell R-500 cartridge (Baxter Healthcare), and late-stage schizonts of P knowlesi were isolated by centrifugation on 45% Percoll (Pharmacia) gradients and cultured 3-4 hr in RPM1 1640 culture medium with 50 uglml chymostatin and 50 uglml leupeptin. The RNA was extracted by a single-step method using 4 M guanidinium thiocyanate, 25 mM sodium citrate (pH 7.0) 0.5% sarcosyl, and 100 mM 8-mercaptoethanol (Chomczynski and Sacchi, 1987) (RNAzol, CinnalBiotecx).
+ 81.5
- (875/[bp
in oligonucleotide]);
“final
wash”
is the final
For Northern blot analysis, the RNA from the fifth preparation was enriched for the poly(A) fraction twice purified over oligo(dT)-cellulose spun columns (Pharmacia). The poly(A)-enriched RNA (I rcg per lane) was separated by agarose gel electrophoresis (1% agarose, 20 mM MOPS, 5 mM sodium acetate, 0.5 mM EDNA, 200 mM formalin. 0.5 &ml ethidium bromide), washed 2 hr in several changes of diethyl pyrocarbonate-treated distilled water, equilibrated in 20x SSC, transferred onto GeneScreen Plus (Du Pont). cross-linked onto the membrane with ultraviolet (Stratagene), and dried under vacuum (modified from standard procedures of Maniatis et al., 1982). Northern blots were hybridized with the plC1 insert and oligonucleotides (111, CTTlGTCTATTGATCATCTTTT; 112, TTAATCTAGCTTCCACTCTAAT; 113, AAGGAATAATGCAGAATGGTGT; 60, GAAGCTCCACAGATATTGAGCACA) using the same procedures described above for Southern blots. In vitro translations of mRNA were performed with rabbit reticulocyte lysate according to manufacturer’s recommendations (Promega Biotech) with 5 ug total RNA. The in vitro translated products were immunoprecipitated as described below. Hybrid arrest of in vitro translation was performed as described previously (Pines and Hunt, 1987) with sense (oligo 30: ATGGGAACTAATATGGA) and antisense (oligo 31: TCCATATTAGTTCCCAT) oligodeoxynucleotides (Synthecell Corporation). Oligodeoxynucleotide (10 ng in 1 trl of distilled water) was combined with 5 ug of total RNA (in 5 ul of 10 mM HEPES), incubated at 3pc for 20 min, and treated with 1 U of RNAase H (in 1 ul of 10 mM Tris [pH 7.51, 50 mM KCI, 5 mM MgClz, 0.05 mM dithiothreitol, 25 ug of bovine serum albumin, 25% glycerol) (Bethesda Research Laboratories) for 30 min at 3pC before in vitro translation as above. Production of Antisera to Fuslon Pmteins and Peptides Nonoverlapping fragments DR.1 and DR.2 were created by polymerase chain reaction (PCR) using oligonucleotide primers derived from the sequence of plC1 (see Figure 1). Each PCR fragment was constructed with a 5’ BamHl restriction site in frame with the pATH2 vector (pJH12) trpE open reading frame (Spindler et al., 1984; provided by T. J. Koerner) and at the 3’end with multiple stop codons ending a Sal1 restriction site (DR.1 5’ oligonucleotide primer: CCGGGGATCCGCAAAAlGAGGGfGCAACTGCG; DR.1 3’ oligonucleotide primer: TTTTGTCGACCCGAACCGTTCATATACTTCTC; DR.2 5’ oligonucleotide primer: GGGGATCCGGGAACTGATGAAAAGGCCAAG; DR.2 3’ oligonucleotide primer: GGGGTCGACTTATTAATTGCCAGATCCAGGAACATT). The PCR was run for 30 cycles of 92OC for 1 min, 45’X for 1 min, and 74% for 4 min (plus 4 s added each cycle). Each PCR product was purified by phenol-chloroform and ethanol precipitation, digested with BamHl and Sall, purified byagarose gel electrophoresis, isolated on glass powder, and ligated to pATH2. E. coli strain DH5a
Cell 152
(Bethesda Research Laboratories) was transformed with pATH2/DR.l or pATH2/DR.2. Liquid cultures of each clone and a wild-type culture were induced to produce trpE fusion proteins by the addition of indole acrylic acid (10 ug/ml) and the cells were grown until saturation. Cells expressing fusion proteins were recovered by centrifugation 20 min at 5000 x g, resuspended in 10 ml of 0.3 M NaCI. 0.5 mM EDTA, and 50 mM Tris (pH 74) (TEN) containing 1 mg/ml lysozyme and incubated for 15 min on ice. After addition of 0.5 ml of 4% Triton X-100 (10 min on ice) the solution was mixed with 12 ml of 1.5 M NaCI, 12 mM MgC& containing 23 ul of 20 mg/ml DNAase and incubated for 60 min at 4%. Insoluble material was pelleted by centrifugation at 4000 x g for 15 min and washed three times in TEN buffer. Triton X-100 insoluble pellets containing fusion proteins of DR.1 (60 kd) and DR.2 (100 kd) or wild-type trpE (45 kd) were separated by SDS-PAGE. The fusion proteins were electroeluted (Bio-Rad) from gel slices into 25 mM Tris (pH 6.3) 192 mM glycine, and 0.1% SDS. The electroeluted fusion proteins were emulsified with Freund’s complete adjuvant for primary immunizations and incomplete adjuvant for each booster immunization. The C-terminal peptide (DIEHMQQFTPLDYS) and peptide 3 (EGKSSTNEADPGSQSGAPASRS) (7.5 ug of each purified by high pressure liquid chromatography) were conjugated to keyhole limpet hemocyanin (7.5 ug) (Calbiochem) overnight at room temperature in 7.5 ml of PBS using 940 pl of 0.05% glutaraldehyde. Conjugated proteins were dialyzed for 24 hr with 500 vol of PBS. Rabbits were immunized with conjugated proteins: primary immunizations were emulsified in Freund’s complete adjuvant and boosting immunizations were emulsified in incomplete adjuvant.
Erythmcyte
Preadeorptions
and Immunoprecipitations
Culture supernatants were incubated with erythrocytes from various hosts to remove the molecules that bound to these erythrocytes. Culture supernatant was incubated twice for 30 min with packed erythrocytes at a ratio of 0.5 ml of packed erythrocytes per ml of culture supernatants (Figure 6) or at a ratio of 1 ml of packed erythrocytes per 1 ml of culture supernatant (Figure 7). lmmunoprecipitations were modified from techniques described previously (David et al., 1964). Parasite extracts were made from 1 x IO9 cultured schizonts extracted in 3 ml of 1% Triton X-100 in 10 mM HEPES (pH 6.5) 50 us/ml chymostatin, and 50 uglml leupeptin and centrifuged for 20 min at 20,000 x g. Culture supernatants and detergent extracts (400 pl each) were incubated for 30 min with 5 ul of serum, then 75 ul of a 50% suspension of protein A-Sepharose CL-46 (Pharmacia) was added and incubated another 30 min, washed once with 05% bovine serum albumin in 0.5% Briton X-100,0.15 M NaCI, 1 mM EDTA, and 50 mM Tris (pH 74) (NETT), once in NETT, once in NETT with 0.5 M NaCI, twice in NETT, and separated by SDS-PAGE. To help separate the closely migrating proteins of interest, SDS-PAGE was performed according to the Laemmli method using 0.6% instead of 06% bisacrylamide in a total of 30.0% acrylamide monomer and a mixed-chain-length SDS preparation (69% lauryl, 24% myristyl, 5% cetyl sulfate salts; Sigma) (Margulies and Tiffany, 1964) in SDS-PAGE electrophoresis buffers. Some experiments used 99.9% pure SDS and/or 0.6% bis as a cross-linker as noted in the figure legends. All SDS-PAGE gels contained a final concentration of 75% acrylamide monomer and were either 1.5 mm or 0.75 mm thick, as indicated in the figure legends.
Pulse-Chase
Analysis
F! knowlesi cultures were incubated for 1 hr with 150 uCi of [sS]methionine/cysteine (ICN Radiochemicals), 50 ug/ml chymostatin, and 50 uglml leupeptin in RPM1 1640 culture medium, washed two times in culture medium, separated into five aliquots (2 x 108 schizonts each), and cultured for 0,90, 160,270, and 360 min or 0,30,60,90, and 540 min. At each time point an aliquot was centrifuged for 5 min at 1000 x g. Supernatants and pellets were separated. Pellets were then either frozen immediately to -7ooC or denatured immediately in boiling SDS. Samples denatured in SDS (Gerace and Blobel, 1960) were boiled 5 min in 0.5% SDS, 50 mM Tris (pH 7.4) 100 mM NaCI, and 2 mM EDTA, cooled, mixed with protease inhibitors (0.5 mM phenylmethylsulfonyl fluoride, 1 mM TLCK, 1 mM TPCK, 50 ug/ml chymostatin, 50 &!/ml leupeptin), and frozen to -7ooC. When used these rapidly denatured samples were boiled again, fresh protease inhibitors were
added along with 10 rig/ml DNAase I and 10,mM Mg&, mixed with Triton X-100 to a final concentration of 2%, centrifuged 5 min at 15,000 x g, and the supernatant was used in immunoprecipitations.
Immunofluoreacence
Yicmscopy
All incubations were at 23oc. Thin films of cultured P knowlesi schizonts in 50% fetal bovine serum on glass slides were air dried and fixed 5 min in PBS containing 1% formaldehyde. The fixed samples were rinsed in PBS and incubated for 5 min in block buffer (PBS containing 0.1% Triton X-100 and 2.6 mg/ml normal goat serum). The slides were then incubated for 60 min in a humidified chamber with immune or nonimmune serum diluted BOO-fold in block buffer, washed with block buffer (three times, 5 min each), and incubated for 30 min with rhodamine-conjugated goat anti-rabbit IgG (Southern Biotechnology Associates) diluted 250-fold in block buffer. Slides were washed with block buffer (three times, 5 min each), mounted in 90% glycerol, 10 mM Tris (pH 7.4) and bisbenzimide H33256 (Calbiochem), and viewed on a Zeiss Axiophot fluorescence microscope.
Immunoelectron
Mlcmscopy
Samples were prepared for immunoelectron microscopy by fixation in 1% formalin and 20 mM ethyl acetimidate HCI (Geiger et al., 1961) (EAI; Serva Feinbiochemica) for 5 min in RPM1 1640 culture medium followed by the addition of either 6% glutaraldehyde to a final concentration of 0.1% for 15 min at room temperature or 100 mM ethylene glycolbis(succinimidylsuccinate) (EGS; Pierce) in DMSC to a final concentration of 10 mM at room temperature or 3pC for 30 min, then washed and stored in RPM1 1640 culture medium and 0.1% NaNa. Other samples were fixed with 1% formalin in PBS only for 1 hr at room temperature, then washed and stored in PBS with NaN3 until embedded. Glutaraldehyde as the only fixative destroyed reactivity with all antisera. All samples were dehydrated, embedded in LR White, and probed with antibody as described previously (Torii et al., 1969).
Ferrttln-Brtdge
Immunoelectmn
Micmscopy
This procedure was modified from that described previously (Willingham, 1960). Specimens were fixed in 1% formalin in PBS for 1 hr at room temperature and washed three times in PBS with 200 mM NaCl (350 mM NaCl total), 0.1% Triton X-100. The buffer in all subsequent steps contained 200 mM NaCI, 1 mM EDTA, and 2.5 mg/ml goat serum (Jackson Laboratories) added to PBS (pH 7.4). Fixed samples were incubated for 1 hr at 4°C in each of the following antibody solutions followed by six quick rinses and a 30 min wash: primary antibody (rabbit sera, anti-DR.1 and anti-DR.2, were each diluted 1:200 in the washing buffer); secondary antibody (2.2 mg/ml affinity-purified goat anti-rabbit IgG [Jackson Laboratories]); tertiary antibody (2.3 mg/ml affinity. purified rabbit anti-horse ferritin [Jackson Laboratories]); electrondense label (200 pg/ml horse spleen ferritin [Sigma]). Samples were then fixed overnight at 4“C in 0.1% glutaraldehyde in PBS containing 200 mM NaCl and 0.1% Triton X-100, washed in PBS containing 0.05% NaN3, postfixed in 1% osmium tetroxide, dehydrated, and embedded in Epon 612.
We would like to thank Palmer Orlandi, Lee Malay, Sanjai Kumar, David Kaslow, Xiangdong Fang, and Mark Willingham for helpful discussions, Kiet Dan Luc and Anna Milosavljevic for technical assistance, T J. Koerner for use of the pATH2 expression vector, Marvin Shapiro for use of DNADRAW, and John Coligan for synthesis of peptides. We acknowledge support from the US Agency for International Development (grant DPE-0453AOO-4O27-00), the US Public Health Service (grant Al-10645) the US Army R&D Command (grant DAMD-1765C5179) and the UNDPMlorld Bank/WHO Special Programme for Research and Training in Tropical Areas. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 16 USC Section 1734 solely to indicate this fact. Received
March
12, 1990; revised
June
21. 1990.
yF;modium
knowlesi
Duffy
Receptor
Family
Aikawa, M., Miller, L. H., Johnson, J., and Rabbege, cyte entry by malarial parasites. A moving junction cyte and parasite. J. Cell Biol. 77, 72-82.
Margulies, M. M., and Tiffany, H. L. (1984). Importance of sodium dodecyl sulfate source to electrophoretic separations of thylakoid polypeptides. Anal. Biochem. 736, 309-313.
J. (1978). Erythrobetween erythro-
Marsh, W. L. (1975). Present status of the blood Crit. Rev. Clin. Lab. Sci. 5, 387-412.
group
system.
CRC
Aley, S. B., Bates, M. D., Tam, J. t?, and Hollingdale, M. R. (1986). Synthetic peptides from the circumsporozoite proteins of Plasmodium faC ciperum and Plesmodium knowlesi recognize the human hepatoma cell line HepGPA16 in vitro. J. Exp. Med. 764, 19151922.
Mason, S. J., Miller, L. H., Shiroshi, T., Dvorak, J. A., and McGinniss, M. H. (1977). The Duffy blood group determinants: their role in the susceptibility of human and animal erythrocytes to Plasmodium knowlesi malaria. Br. J. Haematol. 36, 327-335.
Chomczynski, P, and Sacchi, N. (1987). Single-step method isolation by acid guanidinium thiocyanate-phenol-chloroform tion. Anal. Biochem. 762, 156-159.
of RNA extrac-
Miller, L. H., Mason, S. J., Dvorak, J. A., McGinniss, M. H., and Rothman, I. K. (1975). Erythrocyte receptors for (Plasmodium knowlesi) malaria: Duffy blood group determinants. Science 789, 561-563.
Cochrane, A. H., Uni, S., Maracic, M., Di Giovanni, L., Aikawa, M., and Nussenzweig, Ft. S. (1989). A circumsporozoite-like protein is present in micronemes of mature blood stages of malaria parasites. Exp. Parasitol. 69, 351-356.
Miller, L. H., Mason, S. J., Clyde, D. F., and McGinniss, M. H. (1976). The resistance factor to Plasmodium vitmx in Blacks: the Duffy blood group genotype FyFy. N. Engl. J. Med. 295, 302-304.
Dame, J. B., Williams, J. L., McCutchan, T. F, Weber, J. L., Wirtz, R. A., Hockmeyer, W. T, Malay, W. L., Haynes, J. D., Schneider, I., Roberts, D., Sanders, G. S., Reddy, E. P, Diggs, C. L., and Miller, L. H. (1984). Structure of the gene encoding the immunodominant surface antigen on the sporozoite of the human malaria parasite Plasmodium falcipafum. Science 225, 593-599. Darnell, J., Lodish, H., and Baltimore, D. (1986). Molecular (Freeman, New York: Scientific American Books).
Cell Biology
David, P H., Had1ey.T. J., Aikawa, M., and Miller, L. H. (1984). Processing of a major parasite surface glycoprotein during the ultimate stages of differentiation in Plasmodium knowlesi. Mol. Biochem. Parasitol. II, 267-282. Davis, L. G., Dibner, M. D., and Battey, J. F. (1986). Basic Methods in Molecular Biology (New York: Elsevier Science Publishing Co., Inc.). Dvorak, J. A., Miller, L. H., Whitehouse, W. C., and Shiroishi, T. (1975). Invasion of erythrocytes by malaria merozoites. Science 787,748-750. Fang, X., Kaslow, D. C., Adams, of the Plasmodium vivax Duffy press.
J. H., and Miller, L. H. (1990). Cloning receptor. Mol. Biochem. Parasitol., in
Fine, E., Aikawa, M., Cochrane, A. H., and Nussenzweig, R. S. (1984). Immune-electron microscopic observations on Plasmodium knowlesi sporozoites: localization of protective antigen and its precursor. Am. J. Trop. Med. Hyg. 33, 220-226. Geiger, B., Dutton, A. H., Tokuyasu, K. T., and Singer, S. J. (1981). Immunoelectron microscope studies of membrane-microfilament interactions: distributions of a-actinin, tropomyosin, and vinculin in intestinal epithelial brush border and chicken gizzard smooth muscle cells. J. Cell Biol. 97, 614-626. Gerace, L., and Blobel, reversibly depolyrnerized
G. (1960). The nuclear envelope during mitosis. Cell 79, 277-287.
lamina
is
Hadley, T J., McGinniss, M. H., Klotz, F. W., and Miller, L. H. (1966). Blood group antigens and invasion of erythrocytes by malaria parasites In Red Cell Antigens and Antibodies, G. Garratty, ed. (Arlington, Virginia: American Association of Blood Banks), pp. 17-33. Haynes, J. D., Dalton, J. T. J., Hudson, D. E., and of a Plasmodium knowlesi ligands on erythrocytes.
l?, Klotz, F W., McGinnis% M. H., Hadley, Miller, L. H. (1988). Receptor-like specificity malarial protein that binds to Duffy antigen J. Exp. Med. 767; 1673-1881.
Hudson, D. E., Wellems, T E., and Miller, L. H. (1988). Molecular basis for mutation in a surface protein expressed by malaria parasites, J. Mol. Biol. 203, 707-714. Kobayashi, S., Eden-McCutchan. F., Framson, P, and Bornstein, t? (1986). Partial amino acid sequence of human thrombospondin as determined by analysis of cDNA clones: homology to malarial circumsporozoite proteins. Biochemistry 25, 8418-8425. Lawler, J., and Hynes, R. 0. (1986). The structure of human thrombospondin, an adhesive glycoprotein with multiple calcium-binding sites and homologies with several different proteins. J. Cell Biol. 703, 1635-1648. Maniatis, T., Fritsch. E. F., and Sambrook. J. (1982). Molecular Cloning: A Laboratory Manual (Cold Spring Harbor, New York: Cold Spring Laboratory).
Miller, L. H., Powers, K. G., and Shiroishi, T. (1977). Plasmodium knowlesi: functional immunity and antimerozoite antibodies in rhesus monkeys after repeated infection. Exp. Parasitol. 47, 105-111. Miller, L. H., Aikawa, M., Johnson, J. G., and Shiroishi, T (1979). Interaction between cytochalasin B-treated malarial parasites and erythrocytes. Attachment and junction formation. J. Exp. Med. 749, 172-184. Miller, L. H., Hudson, Plasmodium knowlesi Parasitol. 37, 217-222.
D., and Haynes, J. D. (1988). erythrocyte binding proteins.
Identification of Mol. Biochem.
Nichols, M. E., Rubinstein, P, Barnwell, J., de Cordoba, S. R., and Rosenfield, R. E. (1967). A new human blood group specificity defined by a murine monoclonal antibody. J. Exp. Med. 766, 776-785. Pines, J., and Hunt, T. (1987). Molecular characterization for cyclin from sea urchin eggs. EMBO J. 6, 2987-2995.
of the mRNA
Spindler, K. R., Rosser, D. S., and Berk, A. J. (1984). Analysis of adenovirus transforming proteins from early regions 1A and IB with antisera to inducible fusion antigens produced in Eschefichia co/i. J. Virol. 49, 132-141. Torii, M., Adams, J. H., Miller, L. H., and Aikawa, M. (1989). Release of merozoite dense granules during erythrocyte invasion by Plasmodium knowlesi. Infect. Immun. 57, 3230-3233. Uni, S., Masuda, A., Stewart, M. J., Igarashi, I., Nussenzweig, R., and Aikawa, M. (1987). Ultrastructural localization of the 150/130 Ko antigens in sexual and asexual blood stages of flesmodium falciparum-infected human erythrocytes. Am. J. Trop. Med. Hyg. 35461-488. Weber, J. L. (1988). A review. Exp. Parasitol. 66, 143-170.
Molecular
biology
of malaria
parasites.
Wellems, T E., and Howard, R. J. (1986). Homologous genes encode two distinct histidine-rich proteins in a cloned isolate of Plesmodium falciparum. Proc. Natl. Acad. Sci. USA 83, 6065-6069. Wertheimer, S. P., and Barnwell, J. W. (1989). Plesmodium vivax interaction with the human Duffy blood group glycoprotein: identification of a parasite receptor-like protein. Exp. Parasitol. 69, 340-350. Willingham, M. C. (1980). Electron localization of intracellular antigens tin bridge procedures. Histochem.
microscopic immunocytochemical in cultured cells: the EGS and ferriJ. 72, 419-434.
Young, R. A., and Davis, R. W. (1983). Efficient isolation of genes using antibody probes. Proc. Natl. Acad. Sci. USA 80, 1194-1198. GenBank
Accession
by
Number
The accession numbers for the sequences M37512 for plC1 and M37513 for pEco6.
reported
in this paper
are