Plasmodium vivax

0 downloads 0 Views 954KB Size Report
Guinea Institute of Medical Research, Madang, Papua New Guinea ... These data identify multiple clonal populations of P. vivax in the PNG population and.
JOBNAME: JEP 83#1 96 PAGE: 1 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

EXPERIMENTAL PARASITOLOGY ARTICLE NO. 0044

83, 11–18 (1996)

Plasmodium vivax: Favored Gene Frequencies of the Merozoite Surface Protein-1 and the Multiplicity of Infection in a Malaria Endemic Region KATHLEEN A. KOLAKOVICH,* ANISA SSENGOBA,* KIMBERLY WOJCIK,* TAKAFUMI TSUBOI,*,1 FADWA AL-YAMAN,† MICHAEL ALPERS,† AND JOHN H. ADAMS*,2 *Department of Biological Sciences, University of Notre Dame, Notre Dame, Indiana 46556, U.S.A.; and †Papua New Guinea Institute of Medical Research, Madang, Papua New Guinea KOLAKOVICH, K. A., SSENGOBA, A., WOJCIK, K., TSUBOI, T., AL-YAMAN, F., ALPERS, M., AND ADAMS, J. H. 1996. Plasmodium vivax: Favored gene frequencies of the merozoite surface protein-1 and the multiplicity of infection in a malaria endemic region. Experimental Parasitology 83, 11–18. In this study, we present an analysis of the Plasmodium vivax MSP-1 polymorphic region 5 and identify a new recombinant gene element. In clinical isolates from Papua New Guinea (PNG), the P. vivax MSP-1 gene type was characterized by restriction fragment length polymorphisms and by Southern blot oligonucleotide hybridizations using probes to type-specific sequences. There were three pairs of dimorphic gene elements in the MSP-1 polymorphic region 5; four of the eight potential different combinations of sequence elements for this region have been identified. The center gene segment was the most polymorphic, especially for the glutamine (Q) repeat element with virtually every gene containing a different length of Q repeats, a finding consistent with database sequence information. The frequencies of all of the polymorphic MSP-1 gene elements were approximately equal except for the first segment, which was biased 10:1 for the Type II (Sal-1 type) versus Type I (Belem type) gene segment. In fact, only one combination (I/Q/S) of the genetic elements containing the type I gene segment for polymorphic region 5 was identified, a finding consistent with sequences reported to gene data banks. Considering only the multiplicity of MSP-1 gene types, 38% of the patients were identified as having multiple infections; when correlated with the circumsporozoite protein and the Duffy antigen binding protein gene types, the multiple infection rate increased to 65% of 23 isolates characterized. Increased age was the only clinical parameter that positively correlated with multiclonal infections and there was no other apparent bias or linkage of gene types among the three loci. These data identify multiple clonal populations of P. vivax in the PNG population and potentially a high rate of concurrent infections in clinical cases. The extreme polymorphism of the MSP-1 polymorphic region 5 suggests that frequent recombination occurs within this gene. The bias in frequency for one recombinant gene motif indicates that intrinsic host or parasite factors may engender increased frequency of one genetic element over another. Failure to identify this type of discrete clonal marker as well as reliance on a single marker can mask the true multiclonal nature of an infection and lead to underestimation of the multiplicity of infection. © 1996 Academic Press, Inc. INDEX DESCRIPTORS AND ABBREVIATIONS: Plasmodium vivax; CSP, circumsporozoite protein; DBP, Duffy antigen binding protein; MSP-1, merozoite surface protein-1; PCR, polymerase chain reaction.

was first shown in an animal model (Holder and Freeman 1981). Sequence analysis of MSP-1 genes derived from different Plasmodium species and clones has identified conserved and semiconserved blocks interspersed with polymorphic regions; these polymorphic regions reflect an allelic-type of sequence dimorphism within species (Tanabe et al. 1987; del Portillo et al. 1988, 1991; Lewis 1989; Miller et al. 1993). Variation in the antigenic phenotype of MSP-1, attributed to variation in MSP-1 gene types, may be related to a lack of heterologous

INTRODUCTION Merozoite surface protein-1 (MSP-1) is the immunodominant antigen expressed on the surface of a malaria merozoite. Much of the interest in studying MSP-1 developed from its potential as an asexual-stage vaccine, since its ability to induce a protective immune response 1 Current address: Department of Parasitology, Ehime University School of Medicine, Shigenobu-Cho, Ehime Japan 91-02. 2 To whom correspondence should be addressed.

11 0014-4894/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

JOBNAME: JEP 83#1 96 PAGE: 2 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

12

KOLAKOVICH ET AL.

strain immunity (McBride et al. 1982; Siddiqui et al. 1987); MSP-1 variation may result from intragenic recombination in the polymorphic regions between the dimorphic allelic types (reviewed by (Miller et al. 1993)) or by immune selection (Hughes 1992). The MSP-1 has been the most thoroughly characterized in Plasmodium falciparum. The extensive recombination between P. falciparum gene sequence types is restricted to the 59 end of the gene with polymorphic region 4 as the last region that exhibits recombination between gene types. Since meiotic recombination readily occurs during mosquito transmission (Walliker et al. 1971, 1987; Ranford-Cartwright 1991; Burkot et al. 1992), this was long thought to be the mechanism for generating recombinant gene types. However, asexual-stage gene conversion, an additional mechanism for intragenic recombination, may play a more important role in generating diversity in MSP-1 (Miller et al. 1993). Previous studies of the Plasmodium vivax MSP-1 have provided clear evidence for genetic recombination in polymorphic regions 4 and 5 (Cheng et al. 1993; Premawansa et al. 1993), indicating a similar although different pattern of genetic diversity for this species relative to P. falciparum. Our data extend this information by identifying a new recombination site in the P. vivax MSP-1 polymorphic region 5 and by determining a bias in gene frequency. Papua New Guinea (PNG) is holoendemic for vivax malaria where residents, even at an early age, are continually infected and reinfected through year-round mosquito transmission. Multiclonal infections by malaria parasites increases the probability for genetic recombination during the sexual cycle in mosquitoes, leading to the generation of sporozoites with unique genotypes. Since blood-stage malaria parasites are haploid organisms, the use of single-copy polymorphic genes such as that for the P. vivax merozoite surface protein-1 facilitates identification of multiclonal parasite infections. MATERIALS

AND

and diagnosed by microscopy as a P. vivax infection (Tsuboi et al. 1994). Whole blood samples were frozen at −75°C or the packed cells and plasma were separated and then frozen at −75°C in EDTA tubes. The parasite DNA was extracted from the blood after lysed erythrocytes were washed in TSE [10 mM Tris (pH 8.0), 1 mM EDTA, 10 mM NaCl] and treatment with proteinase K using phenol:chloroform:isoamyl alcohol (25:24:1) or with guanidine thiocyanate and a nuclease-binding matrix. MSP-1 gene amplification. The MSP-1 polymorphic region 5 was amplified using the polymerase chain reaction (PCR) with primers that annealed in the conserved flanking regions [Table I; based upon Premawansa et al. (1993)]. The reaction mixture consisted of 50 mM KCl, 10 mM Tris–HCl (pH 9.0 at 25°C), 0.1% Triton X-100, 1.5 mM MgCl2, 0.2 mM dNTPs, 600 ng of each primer, ≈150 ng DNA, and 2.5 U Taq DNA polymerase to a final volume of 100 ml and sealed against evaporation with light mineral oil. MSP-1 genes were amplified by 35 cycles of 94°C for 30 sec, 60°C for 1 min, and 74°C for 1 min. Analysis of MSP-1 sequences. The PCR product was digested directly with RsaI, separated by agarose gel electrophoresis, denatured (0.4 N NaOH, 1.5 M NaCl), and blotted onto Hybond N+ (Amersham) nylon membrane. The blots were probed with [g-32p]ATP-labeled oligonucleotides probes complementary to reported sequences of the polymorphic region 5 of the P. vivax MSP-1 (Table I). MSP-1 gene types were detected by autoradiography of the hybridized blots to each probe after a final wash of maximum stringency in 6× SSC, 0.5% SDS. Nucleotide sequences were determined by the dideoxynucleotide chain termination method either from the PCR products or from gene fragments cloned into a plasmid pT7 (Novagen). Direct sequencing of the PCR products was carried out with a Taq DNA polymerase (BRL) and sequencing of the plasmid was performed with a T7 DNA polymerase (US Biochemical). Nucleotide sequence alignments were done with ALIGNMENT (Geneworks 2.1, Intelligenetics) and were manually adjusted to give the best fit. Identification of CSP and DBP gene types. Conditions for the PCR amplification were adapted from methods described previously for the CSP (Qari et al. 1992) and Duffy antigen binding protein genes (DBP) (Tsuboi et al. 1994) using primers and probes described in Table I. The PCR product DNA was separated by agarose gel electrophoresis, denatured, and blotted onto Hybond N+ nylon paper as above. The blots of the PCR-amplified CSP gene fragments were probed with radiolabeled oligonucleotides complementary to the two different tandem repeat types, VK210 and VK247 (Rosenberg 1989; Qari et al. 1992). The blots of the PCR-amplified DBP gene fragments were probed with radiolabeled oligonucleotides complementary to an internal sequence polymorphism associated with a recombination site (Tsuboi et al. 1994).

METHODS

Parasite collection and DNA preparation. Blood was collected from patients presenting acute symptomatic malaria

RESULTS The polymorphic region 5 of P. vivax was

JOBNAME: JEP 83#1 96 PAGE: 3 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

Note. The source for the design of the oligonucleotide primers and probes were: CSP (Qari et al. 1992), MSP-1 (present study and Premawansa et al. 1993), and DBP (Tsuboi et al. 1994).

59-CTTCAAATTCCTTTTTCATG-39 59-TTCCGCAG(G/T)(C/T)CCAT(C/T)GCT-39 59-TTCGTAGATTC(C/T)GCAAA(C/T)TCC-39 59-GAAGATATCAATTATGTATG-39 1 2 DBP

Sal-1 PNG

59-GGGAATTCTTGTGACATGTCGTAAGCG-39 59-CACAACCAATGCGGTAACATC-39 59-CAATACAGTCAATGCGCCAAAC-39 59-CATCAAGTAG(T/C)AAATGCAGTAACG-39 59-GTACAACAACAACAACAACAACAAC-39 59-TTGGTGTTGAGGCTACCTGTC-39 59-TTGGTGCTGGTGTGGCTGATAAC-39 59-GGGAATTCTACTACTTGATGGTCCTC-39 1 2 P Q R S MSP-1

Belem Sal-1 Sal-1 Belem Sal-1 Belem

59-CAGCGGATCCTTAATTGAATAATGCTAGG-39 59-GTCGGAATTCATGAAGAACTTCATTCTC-39 59-CAGCGGATCCTTAATTGAATAATGCTAGG-39 59-GTCGGAATTCATGAAGAACTTCATTCTC-39 1 2 CSP

VK210 VK242

Probe 59 PCR oligonucleotide Original clone Type Gene

TABLE I Oligonucleotide Sequences for PCR Primers and Hybridization Probes

39 PCR oligonucleotide

P. vivax: GENE FREQUENCIES AND MULTIPLICITY OF INFECTION

13

PCR amplified using primers to the flanking conserved sequence in 40 of 50 clinical isolates from Papua New Guinea. In order to identify the gene type, the PCR products were digested with the restriction of endonuclease RsaI and probed with oligonucleotides to internal regions of the restriction fragments (Fig. 1, Table I). The RsaI cleaved at or near the boundaries between the three internal recombinant elements of the polymorphic region 5 (Fig. 2), facilitating identification of size polymorphisms present in the separate gene elements. Relying only on the size of the intact polymorphic region 5 PCR product was not an accurate indicator of gene type. An extreme bias (>10:1) was identified in the PNG samples for MSP-1 genes carrying the Type II sequence (Fig. 1) versus genes with the Type I sequence. No size polymorphism was detected for either Type I or II gene elements. No bias was identified for the second and third recombinant elements of MSP-1 polymorphic region 5 as indicated by random frequency for the gene types containing these elements. The second element had either a block of Q repeats (Belem-like) or a P-type sequence (Sal-1-like). Oligonucleotide hybridization of the RsaIdigested MSP-1 gene fragment, using a probe to the Q repeats, identified this segment as extremely size polymorphic (35–85 bp) such that each of the PNG samples appeared to have a variable number of Q repeats (Figs. 1 and 3). These data for the MSP-1 Q repeats from the PNG samples are consistent with that found in Sri Lanka and the sequences submitted to date to gene databases, since each publicly available P. vivax MSP-1 sequence has a different number of Q repeats (when present). The P-type sequence of the second gene element was much less polymorphic, having a common nucleic acid mutation only at its 59 end, near the recombination site, adding or deleting a Q codon. Recombination between the first and second gene elements of the MSP-1 polymorphic region 5 may be facilitated by the conserved nucleic acid sequence GTACAAC at the recombination site for both the P and Q gene types. The difference in size between the P and Q gene elements

JOBNAME: JEP 83#1 96 PAGE: 4 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

14

KOLAKOVICH ET AL.

FIG. 1. Southern blot hybridization of RsaI-digested P. vivax MSP-1 polymorphic region 5. The polymorphic region 5 was PCR-amplified using primers to the conserved flanking regions, restricted with RsaI, and hybridized with oligonucleotides specific to the different genetic elements within this region. Multiclonal infections were identified by restriction fragment length polymorphisms for a genetic element or when both sequence types were present for a given gene segment (I or II, P or Q, R or S). Differences in signal intensity seem to be due to differences in the amount of PCR product present for the samples and not a result of differences in probe:template hybridization efficiency.

(≈100 bp vs ≈54 bp, respectively) appears to be the main contributing element for the observed size dimorphism reported for this polymorphic region (Porto et al. 1992; Premawansa et al. 1993). A new recombination site was identified in the 39 half of the P. vivax MSP-1 polymorphic region 5 (Figs. 2 and 3). This recombination site

creates the possibility of eight potential combinations of gene elements for the polymorphic region 5; however, so far only four of these have been found in clinical isolates or laboratory strains of P. vivax. The lack of identification of some of the combinations may be due to the relatively few clinical samples examined or an inherent bias against this genetic element.

JOBNAME: JEP 83#1 96 PAGE: 5 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

P. vivax: GENE FREQUENCIES AND MULTIPLICITY OF INFECTION

FIG. 2. Schematic diagram of the P. vivax MSP-1 showing the possible alternate combinations of conserved and semiconserved blocks within the polymorphic region 5. Isolates or clones identified for a specific combination of genetic elements for polymorphic region 5 are shown at the left.

15

(Tsuboi et al. 1994), the multiple infection rate increased to 65% of 23 characterized for all three loci (Table II). Several cases had single infections according to one gene type, but were demonstrated to be multiple infections according to another. Increased age was the only clinical parameter that positively correlated with multiclonal infections and the oldest singly infected individual was a child 3.5 years of age (Table III). The ages of the patients spanned from 3 months to over 12 years with a mean of 3.2 years and a median of 2.5 years. There was no other apparent bias or linkage of gene types among the three loci. A high rate of genetic exchange is evidenced by nearly all possible combinations represented in single infections (Table IV). None of the gene types for any of the three genes showed a particular association with a given gene type of any of the other two genes. These data identify multiple clonal populations of P. vivax in the PNG population, indicate that substantial genetic recombination has occurred in the parasite population, and reveal that a potentially high rate of concurrent infections occurs in clinical cases. DISCUSSION

Using only the multiplicity of MSP-1 gene types, 38% of the patients were identified as having multiple infections; when correlated with the CSP (Fig. 4) and the DBP gene types

The use of single-copy polymorphic genes, such as that for the P. vivax MSP-1, facilitates identification of multiclonal parasite infections

FIG. 3. Recombination elements of the P. vivax MSP-I polymorphic region 5. The deduced amino acid sequence of clones isolated from the PNG isolates is compared to those gene sequences previously reported. Dashed lines as spacers were inserted to the best alignments. Above and below the sequence blocks are the type designations. Clone PNG 38A has the newly identified recombinant type (IQS). This information confirmed the data obtained by Southern blot hybridizations.

JOBNAME: JEP 83#1 96 PAGE: 6 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

16

KOLAKOVICH ET AL. TABLE III Multiplicity of Plasmodium vivax Infections Related to Age

FIG. 4. An example of the identification of P. vivax CSP central repeat region by Southern blot hybridization with type-specific oligonucleotide probes (Table I (Rosenberg, 1989; Qari et al., 1992, 1994)). Probes to the two characterized P. vivax CSP repeat types were both 27 bases in length: the type 1 repeat translated as GDRADGQPA and the type 2 repeat encoded ANGAGNQPG.

Number of patients Age span (in months) Average age (in months) a

Single infections

Multiple infections

8 4–44 26

15 9–132 44a

No age available for two patients.

from haploid blood-stage malaria parasites. In this study of P. vivax clinical isolates from Papua New Guinea, we present an analysis of the MSP-1 polymorphic region 5 along with data about the CSP and DBP gene polymorphisms to assess the multiplicity of vivax malaria infections. Previous studies have identified two predominant sequence types for this polymorphic region based on the MSP-1 genes sequenced from the P. vivax laboratory strains Belem and Sal-1 (del Portillo et al. 1991; Porto et al. 1992). Analysis of P. vivax isolates from Sri Lanka identified recombination between these types to create a third gene type. This identified a recombinant gene element at the beginning of the polymorphic region 5 termed blocks I (Belem) and II (Sal-1) (Premawansa et al. 1993). The recombinant Sri Lanka MSP-1 genes had the type II Sal-1 block immediately followed by variable numbers of glutamine repeats and fol-

lowed by another Belem-like sequence. The data presented here identify the sequence following the Q repeats (and the P element) as an additional recombinant element of the polymorphic region 5 and indicate that double recombinations may occur. The extreme polymorphism of the MSP-1 polymorphic region 5 indicates that frequent recombination occurs within this gene. Nevertheless, the bias in frequency for one recombinant gene motif suggests that intrinsic host or parasite factors may engender increased frequency of one genetic element over another. Failure to identify this type of discrete clonal marker as well as reliance on a single marker can mask the true multiclonal nature of an infection and lead to underestimation of the multiplicity of infection. Multiclonal infections by malaria parasites increases the probability for heterogametic genetic recombination during the sexual cycle in mosquitoes, leading to the generation of sporozoites with unique genotypes. This study provides positive evidence for ge-

TABLE II Multiplicity of Infection in Vivax Malaria Patients

TABLE IV Genotypes of Single Plasmodium vivax Infections

Genes

Percentage with multiple alleles detected

CSP MSP-1 DBP

32 38 23

CSP + MSP-1 CSP + DBP MSP-1 + DBP

58 48 43

CSP + MSP-1 + DBP

65

P#

CSP

MSP-1

DBP

Age (months)

46 40 38 8 10 42a 31a 6

1 1 2 1 2 1 2 2

2PR 2PR 2PS 1QS 2PR 2QS 2QS 2QS

1 1 2 1 2 2 1 1-2 Hybrid

4 10 22 26 24 36 42 44

a

Concurrent infection with P. falciparum.

JOBNAME: JEP 83#1 96 PAGE: 7 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

P. vivax: GENE FREQUENCIES AND MULTIPLICITY OF INFECTION netic recombination in the P. vivax population of PNG. Even though only three different loci were examined in this small study group, our data suggest that humans in this area are being repeatedly infected with distinct parasites. Patients were infected with multiple parasite genotypes and the various genotypes observed indicate a high level of genetic recombination. Genetic recombination is advantageous for the parasite population because it continually leads to infections of new antigenic types combining new and old polymorphic gene types. Genetic recombination coupled with multiple concurrent infections and the high transmission rates may help perpetuate malaria in the PNG population. ACKNOWLEDGMENTS We thank Drs. Lal and Qari for providing CSP probes and primers used in this study and Karen Kutz for her comments on this paper. This work was supported in part by Public Health Service Grant R29 AI33656 from the National Institute of Allergy and Infectious Diseases and a Faculty Research Project Grant from the University of Notre Dame. K. Kolakovich and A. Ssengoba were supported by summer research fellowships from the Howard Hughes Medical Institute.

REFERENCES BURKOT, T. R., WIRTZ, R. A., PARU, R., GARNER, P., AND ALPERS, M. P. 1992. The population dynamics in mosquitoes and humans of two Plasmodium vivax polymorphs distinguished by different circumsporozoite protein repeat regions. American Journal of Tropical Medicine and Hygiene 47, 778–786. CHENG, Q., STOWERS, A., HUANG, T. Y., BUSTOS, D., HUANG, Y. M., RZEPCZYK, C., AND SAUL, A. 1993. Polymorphism in Plasmodium vivax msa1 gene—the result of intragenic recombinations. Parasitology 106, 335–345. del PORTILLO, H. A., GYSIN, J., MATTEI, D. M., KHOURI, E., UDAGAMA, P. V., MENDIS, K. N., AND DAVID, P. H. 1988. Plasmodium vivax: Cloning and expression of a major blood-stage surface antigen. Experimental Parasiology 67, 346–353. del PORTILLO, H. A., LONGACRE, S., KHOURI, E., AND DAVID, P. H. 1991. Primary structure of the merozoite surface antigen 1 of Plasmodium vivax reveals sequences conserved between different Plasmodium species. Proceedings of the National Academy of Sciences of the United States of America 88, 4030–4034. HOLDER, A. A., AND FREEMAN, R. R. 1981. Immunization against blood-stage rodent malaria using purified parasite antigens. Nature 294, 361–364.

17

HUGHES, A. L. 1992. Positive selection and interallelic recombination at the merozoite surface antigen-1 (MSA-1) locus of Plasmodium falciparum. Molecular Biology and Evolution 9, 381–393. LEWIS, A. P. 1989. Cloning and analysis of the gene encoding the 230-kilodalton merozoite surface antigen. Molecular and Biochemical Parasitology 36, 271. MCBRIDE, J. S., Walliker, D., AND MORGAN, G. 1982. Antigenic diversity in the human malaria parasite Plasmodium falciparum. Science 217, 254–257. MILLER, L. H., ROBERTS, T., SHAHABUDDIN, M., AND MCCUTCHAN, T. F. 1993. Analysis of sequence diversity in the Plasmodium falciparum merozoite surface protein-1 (MSP-1). Molecular and Biochemical Parasitology 59, 1–14. PORTO, M., FERREIRA, M. U., CAMARGO, L. M. A., PREMAWANSA, S., AND del PORTILLO, H. A. 1992. Second form in a segment of the merozoite surface protein-1 gene of Plasmodium vivax among isolates from rondonia (Brazil)—short communication. Molecular and Biochemical Parasitology 54, 121–124. PREMAWANSA, S., SNEWIN, V. A., KHOURI, E., MENDIS, K. N., AND DAVID, P. H. 1993. Plasmodium vivax: Recombination between potential allelic types of the merozoite surface protein msp1 in parasites isolated from patients. Experimental Parasitology 76, 192–199. QARI, S. H., COLLINS, W. E., LOBEL, H. O., TAYLOR, F., AND LAL, A. A. 1994. A study of polymorphism in the circumsporozoite protein of human malaria parasites. American Journal of Tropical Medicine and Hygiene 50, 45–51. QARI, S. H., GOLDMAN, I. F., POVOA, M. M., DISANTI, S., ALPERS, M. P., AND LAL, A. A., 1992. Polymorphism in the circumsporozoite protein of the human malaria parasite Plasmodium vivax.. Molecular and Biochemical Parasitology 55, 105–113. RANFORD-CARTWRIGHT, L. C., BALFE, P., CARTER, R., AND WALLIKER, D. 1991. Genetic hybrids of Plasmodium falciparum identified by amplification of genomic DNA from single oocysts. Molecular and Biochemical Parasitology 49, 239–244. ROSENBERG, R., W IRTZ , R. A., L ANAR , D. E., S ATTA BONGKOT, J., HALL, T., WATERS, A. P., AND PRASITTISUK, C. 1989. Circumsporozoite protein heterogeneity in the human malaria parasite Plasmodium vivax. Science 245, 973–976. SIDDIQUI, W. A., TAM, L. Q., KRAMER, K. J., HUI, G. S., CASE, S. E., YAMAGA, K. M., CHANG, S. P., CHAN, E. B., AND KAN, S. C. 1987. Merozoite surface coat precursor protein completely protects Aotus monkeys against Plasmodium falciparum malaria. Proceedings of the National Academy of Sciences of the United States of America 84, 3014–3018. TANABE, K., MACKAY, M., GORMAN, M., AND SCAIFE, J. G. 1987. Allelic dimorphism in a surface antigen gene of the

JOBNAME: JEP 83#1 96 PAGE: 8 SESS: 21 OUTPUT: Thu Jun 20 15:44:29 1996 /xypage/worksmart/tsp000/70152f/7

18

KOLAKOVICH ET AL.

malaria parasite Plasmodium falciparum. Journal of Molecular Biology 195, 273–287. TSUBOI, T., KAPPE, S. H. I., AL-YAMAN, F., PRICKETT, M. D., ALPERS, M., AND ADAMS, J. H. 1994. Natural variation within the principal adhesion domain of the Plasmodium vivax Duffy binding protein. Infection and Immunity 62, 5581–5586. WALLIKER, D., CARTER, R., AND MORGAN, S. 1971. Genetic recombination in malaria parasites. Nature 232, 561–562.

W ALLIKER , D., Q UAKYI , I. A., W ELLEMS , T. E., M C CUTCHAN, T. F., SZARFMAN, A., LONDON, W. T., CORCORAN, L. M., BURKOT, T. R., AND CARTER, R. 1987. Genetic analysis of the human malaria parasite Plasmodium falciparum. Science 236, 1661–1666.

Received 4 December 1995; accepted 6 February 1996