| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(msp3). We found a median of 4 (range = 2–6) and 12 (range = 2–23) unique genotypes based on dbpII and msp3, respectively, per person at one time point and at least 12–33 unique genotypes per person over a four-month period. Control polymerase chain reactions (PCRs) detected 0–31% of clones with haplotypes that arose from PCR artifacts, indicating that caution must be taken when using PCR-based analysis to examine complex infections. To reduce artifacts from clones, analysis was based on haplotypes unlikely to have been generated by PCR artifacts or had been previously identified. Plasmodium vivax infections can be highly complex in disease-endemic areas, suggesting continual genetic mixing that could have significant implications for the use of antimalarial drugs and malaria vaccines. |
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plasmodium parasites are haploid for most of their life cycle except for a brief period in the mosquito vector in which sexual recombination occurs as an obligate part of the parasite life cycle. For sexual recombination to occur, more than one simultaneously infecting parasite genotype per human host must be present. It is possible that interrupted feeding by the anopheline vector and consequent sampling of multiple hosts might also contribute to the diversity of parasites in a blood meal;3 however, this contribution is considered negligible.4,5 In Papua New Guinea, which is endemic for all four species of human malaria, the number of unique genotypes of P. falciparum per individual was estimated to be 1.4–1.9 based on identification of polymorphic loci.6,7 This relatively low multiplicity of P. falciparum infections typically occurs in regions of low or seasonal transmission, such as in the Sudan, where only 20% of P. falciparum–positive samples had multiple infections.8 In areas with continuous and intense transmission such as in Nigeria, the proportion of individuals with multiple parasite infections increases to 94% of P. falciparum–positive samples.9 In Tanzania, which is one of the most intense areas of P. falciparum malaria transmission in the world, 10-fold greater than that observed in Papua New Guinea, an average of 3–5 clones were found to simultaneously infect an individual.10,11 The greatest number of unique P. falciparum clones reported in a human host at a single time point is 12.12 Overall, the multiplicity of P. falciparum is weakly proportional to transmission intensity but only for very large differences of 10–100 fold.13,14 What limits the number of clones that simultaneously infect an individual likely involves a complex set of factors such as inbreeding,7 but it is generally agreed that the complexity of infection based on analysis of direct polymerase chain reaction (PCR) products of polymorphic genes represents a minimal estimate.14,15
Only a few studies have examined the multiplicity of P. vivax infections. Estimates of the proportion of individuals with multiple P. vivax infections have ranged from 30% to 36% in Thailand16,17 and 33% to 60% in India18 to more than 65% in Papua New Guinea6,19,20 in populations with very different transmission intensities. The average number of unique P. vivax clones simultaneously infecting a single individual is generally reported to be less than two, and the maximum number reported is six based on restriction fragment length polymorphism (RFLP) analysis of the merozoite surface protein 3 (msp3).6 Although these estimates of diversity can not be directly compared because of differences in sampling and genotyping methods, they likely represent significant underestimates of multiple clonal infections in an individual. Most studies that have examined clonal diversity of malaria have used a nested PCR to detect multiple infections by RFLP, size polymorphisms, allele-specific primers, or allele-specific probes. This approach could miss less abundant parasite strains. Although previous studies have estimated the complexity of parasite infections from samples collected at one time point, a single P. vivax infection usually persists from three to six months.21 Furthermore, the relative abundance of different malaria clones can vary greatly over time.6,22 Therefore, repeated sampling of the same individual over the span of several months may also increase the ability to detect less abundant parasites. Recently, we sequenced up to 10 clones of the P. vivax Duffy binding protein region II (dbpII) obtained from a single sample and found 6–8 unique clones in each individual, suggesting that examination of a greater number of clones would increase the ability to detect less abundant strains.19 To better estimate the diversity and multiplicity of P. vivax in a holoendemic population in Papua New Guinea, we examined multiple clones from samples collected repeatedly over a four-month period from asymptomatic children and one adult using the highly polymorphic loci that encode the single-copy P. vivax dbpII and msp3 genes.19
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
at the first time point, were randomly selected for detailed study at the additional time points. Informed consent was obtained from all human adult participants and from the parents or legal guardians of children. Study protocols and consent forms were reviewed and approved by institutional review boards at the Veterans Affairs Research Service and Papua New Guinea Institute for Medical Research as part of a larger study examining the prevalence and intensity of P. vivax infection, as previously described.19 Parasitologic diagnosis of malaria. Thick and thin blood smears were stained with 4% Giemsa and examined under an oil-immersion lens for parasitemia as described.19
Polymerase chain reaction amplification and cloning of msp3 and dbpII.
The dbpII and msp3 PCRs with Taq polymerase were conducted as described with 35 cycles during round I and 30 cycles during round II of the PCR.19 The dbpII reactions with Pfu polymerase (Promega, Madison, WI) were similar, except extension times were increased to three minutes for both rounds. The msp3 reactions were also conducted as described, except a that a 16:1 mixture of Taq to Pfu was used to increase the fidelity and yield. As described,19 the dbpII and msp3 PCR products were purified with the Qiaquick Gel Extraction Kit (Qiagen) and cloned directly into either the pCR2.1-TOPO or the pCR-BluntII-TOPO cloning vectors (Invitrogen, Carlsbad, CA) depending on whether Taq or Pfu polymerase was used, respectively. For dbpII, only round II products were cloned because not enough product was generated after one round. For msp3, round I products were cloned whenever enough product was observed; for all other msp3-positive reactions, round II products were cloned.
Identification of P. vivax dbpII genotypes. The presence of one or more dbpII genotypes was determined both by RFLP analysis of sample PCR product and by sequencing 8–19 cloned dbpII inserts for each sample. The RFLP analysis was performed as described on six polymorphic residues that either created or eliminated a restriction enzyme site.19 Briefly, Pst I (K325K), Afl II (L333F), Fsp I (R387R), Hae III (K386N), and Afl III (S447K) (New England BioLabs, Beverly, MA) cut the amplicons identical to the Sal I sequence once within their respective restriction sites. However, each of these sites was eliminated by a mutation creating a different restriction pattern. SfaN I (New England BioLabs) had three restriction sites in the Sal 1 sequence, and a fourth was created by a G to A mutation when the single nucleotide polymorphism occurred at amino acid 390, which created a different restriction pattern. Since the maximum number of mixed infections that could be identified by this RFLP analysis was two, we cloned dbpII amplicons as described earlier in this report and sequenced 8–19 clones from each sample to gain a better understanding of the number of unique dbpII genotypes per infection. Clones were prepared with a commercial kit (Qiagen) and sequenced by vector-based extended M13 forward and reverse primers. DNA sequencing was done at Case Western Reserve University (Cleveland, OH) or MWG Biotech (High Point, NC) by fluorescence-based methodologies using either a 377 or 370 automated DNA sequencer, respectively (Applied Biosystems, Foster City, CA). The sequence alignments were analyzed by using sequence alignment editor BioEdit version 4.8.823 or Sequencher version 4.1.4 (Gene Codes Corporation, Ann Arbor, MI). To reduce the effects of PCR artifacts, only polymorphic residues that occurred in two or more blood samples that had been identified previously or that were unlikely haplotypes due to in vitro artifacts were included. GenBank accession numbers are as follows: AF469523-AF469532, AF469551-AF469582, and AY970837-AY970925.
Identification of P. vivax msp3 genotypes.
The presence of one or more msp3 genotypes was determined both by RFLP analysis of sample PCR product and by the same RFLP analysis on 7–38 cloned msp3 inserts for each sample. Both sample PCR and cloned PCR products were amplified with Taq polymerase as described earlier in this report and digested with Alu I and Hha I, and visualized by gel electrophoresis as described (see Figures 1
and 2 for examples).24
|
|
.
To evaluate the generation of PCR-generated artifacts with our cloning strategy, we mimicked multiple infections by mixing three dbpII or three msp3 clones that had different sequences or RFLP patterns. For the dbpII control reactions, equal amounts of each of the following plasmids were mixed: pEGFPDBP7.18, pEGFPDBP27.16, and pEGFPDBPSal-1.25 A total of 7.5 ng of the mixed dbpII plasmids were then used as template for the nested PCR as described earlier in this report using Pfu polymerase. Since not enough PCR product was generated from the clinical samples for cloning, only round II control products were cloned for further analysis. Sixteen clones were sequenced as described earlier in this report.
For the msp3 control reactions, equal amounts of each of the following plasmids cloned from round one PCR products were mixed: pCRMsp54-25, pCRMsp1094-1, and pCRMsp1094-2, which represented three different restriction patterns and two different-sized PCR products. Round I PCR products amplified with Taq polymerase were generated with 0.5 ng of total mixed msp3 plasmids and cloned and analyzed as described earlier in this report. Round II PCR products were amplified with both Taq and a 16:1 mixture of Taq and Pfu polymerases starting with 0.05 ng of total mixed msp3 plasmids; these products were cloned and analyzed by RFLP as described.24
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
demonstrated a P. vivax prevalence three-fold greater than as determined by blood smear analysis.19
Multiple P. vivax infections simultaneously in children.
Figure 3 shows the complexity of infection as determined by RFLP analysis of the dbpII and msp3 of amplicons directly amplified from samples compared with that obtained by examination of multiple clones. The RFLP analysis of samples for dbpII could only determine if there were 
2 dbpII genotypes. The use of size polymorphisms combined with RFLP analysis of msp3 products directly amplified from the sample enabled identification of at least three P. vivax msp3 genotypes (Figure 1
). For many of the samples, however, the multiple infections resulted in very complex banding patterns, which made deciphering the precise number of unique clones difficult (Figure 1). Thus, by cloning and analyzing multiple clones from each sample for both dbpII and msp3, we were able to identify a greater number of P. vivax genotypes and gain a better estimate of the complexity of infection. For example, lane 8 in Figure 1 shows three PCR products and complex banding patterns for msp3 for both Alu I and Hha I, indicating at least three P. vivax infections. Figure 2 illustrates the same RFLP analysis on 19 clones from this same sample for which we were able to identify at least 10 unique P. vivax infections.
|
shows the number of unique P. vivax clones simultaneously infecting four children and one adult assessed over a four-month period based on analysis of 8–38 clones for the dbpII and the msp3 genes. The number of unique genotypes estimated by dbpII clones ranged from 0 to 6 with a mean of 3.5 genotypes per infected sample at any time point based on sequencing an average of 10 dbpII clones (range = 8–19 clones) per sample. Repeated sampling of individuals over a four-month period identified additional unique genotypes. For two individuals (Figure 3A and B), parasite clones were unique at each sampling point, whereas some of the same dbpII haplotypes were observed in multiple longitudinal samples in the others (Figure 3C–E). Together, each individual had 5–13 unique dbpII genotypes over the four-month sampling period.
The number of parasite strains examined by RFLP analysis of msp3 clones produced an even higher estimate for the number of unique P. vivax clones simultaneously infecting an individual. More clones were examined for msp3 polymorphisms (median = 21 clones, range = 7–38) compared with that for dbpII. The proportion of unique clones compared with the total number of clones examined for msp3 and dbpII were similar, 0.42 and 0.35, respectively. Multiple clones were not generated for four samples for msp3, but different parasite clones could be determined by known size polymorphisms and RFLP analysis of initial PCR products (Figure 3). The median number of unique parasite genotypes simultaneously infecting an individual at one time point based on msp3 was 12 (range = 2–24, Figure 3). The same RFLP pattern was occasionally identified from the same individual with subsequent samples during the four-month period, and a few RFLP patterns were identified in multiple children. Some of the RFLP patterns matched those reported previously in another village in the Madang Province in Papua New Guinea.6 Most RFLP patterns, however, were unique to that individual. Overall, 12–33 total unique msp3 genotypes were observed for each child based on repeated sampling during the four-month period. In contrast, the adult had two unique genotypes.
Relationship of peripheral blood parasitemia with complexity of infection.
Table 1 shows presence and number of P. vivax and P. falciparum parasites per microliter of blood determined by microscopy of blood smears from the same samples from which the dbpII and msp3 genes were cloned. There was no relationship between whether malaria parasites were present by blood smear, either for P. vivax or P. falciparum, and the complexity of infection for these five samples. Indeed, the sample that showed the highest complexity of infection, the sample collected from the five-year-old child in June, was smear negative for P. vivax (Figure 3A and Table 1).
|
inserts. As shown in Table 2, PCR amplicons cloned after the second round of the nested PCR amplification generated chimeras of two of the parental strains in 8–31% of the clones examined. For the dbpII reaction using Pfu polymerase, an artifact was defined as a haplotype that was different from the parental strains used in the mixture. The points of intersection occurred in three different locations for the five chimeras observed: 1) between amino acids 308 and 384 (three chimeras), 2) between amino acids 417 and 424 (one chimera), and 3) between amino acids 390 and 417 (one chimera). The msp3 PCR also generated RFLP patterns that differed from the original cloned strains after the nested PCR (Table 2) in 8% of clones generated with Taq polymerase and 18% of clones generated with a Taq:Pfu polymerase mixture. No artifacts were detected after the first round of the msp3 PCR. Insufficient numbers of amplicons were generated for dbpII after the first round of PCR to clone from any of the field samples collected.
|
clones because they were identified by RFLP analysis. Therefore, we generated clones from round I PCR products for msp3 where PCR-generated artifacts were less likely to occur under these reaction conditions (Table 2). This was performed for six samples, at one time point each for the 5-, 6-, 12-, and 31-year-old individuals, and at two time points for the 8-year-old old individual (Table 3). The proportion of unique RFLP patterns based on the total number of clones examined after the round I PCR (24 unique clones of 54 round I clones examined) was the same as that generated from the round II PCR (64 unique clones of the 144 round II clones, Table 3 and Figure 3). This indicated that formation of PCR artifacts in vitro is unlikely to significantly overestimate the number of unique RFLP patterns identified for msp3.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This higher estimate of multiple infections compared with previous studies resulted from the examination of many clones from a single sample, up to 38 clones examined in one sample. Most studies of parasite complexity examine PCR products generated directly from blood samples and do not clone. This would favor sampling abundant clones and potentially miss clones with lower levels of parasitemia at the time of sampling, but this may also be less subject to PCR artifact formation. The generation and analysis of multiple clones from a single sample is not easily suited to the extensive study of malaria parasite populations, which is the reason we examined only five subjects in the current study. The examination of multiple clones may still underestimate parasite diversity because more unique genotypes were detected with increasing number of clones examined from each individual.
Whereas only one adult was examined in detail in this study, two other adults from this same population at the initial time point were examined previously by sequencing 7–8 dbpII clones.19 Both of the adults examined previously support the data from this study; only one or two different dbpII sequences were identified in each adult, suggesting that the complexity of infections are lower in adults than children.19 Our previous RFLP and PCR size polymorphism analysis of dbpII and msp3 alleles in the same population (358 of 1,025 P. vivax–positive individuals by PCR) indicate that the proportion of individuals with 2 P. vivax infections was highest among 5–9-year-old children and subsequently decreased with age.19 Other studies have shown that the complexity of P. falciparum decrease with age, presumably because adults have developed broader immunity than young children.9,12,13,29,30
We observed, as have others, that PCR errors can generate variants in vitro that may overestimate real parasite diversity.26,31 This may occur by several potential mechanisms. Single basepair mutations can arise from polymerase error. Mutation rates for DNA polymerases are well known, ranging from 1.3 x 10–6 errors per base pair per cycle for Pfu polymerase to 3.0 x 10–5 errors per base pair per cycle with Taq polymerase.32 Thus, we would expect approximately 6% of the dbpII sequences (712 basepairs) examined and between 9% and 16% of msp3 amplicons (1,100–1,900 basepairs) to contain point mutations because we used primarily Pfu polymerase. The PCR errors can also result in apparently novel genotypes through formation of chimeras and heteroduplexes, which the mutation rates do not necessarily take into account. By examining dbpII sequences, we found more single point mutations due to Taq polymerase error than to Pfu polymerase error. In contrast, we found that Pfu polymerase–generated clones had a higher number of chimerical sequences that are likely due to the lower processivity of the Pfu enzyme because chimeric molecules are mainly caused by incompletely extended PCR products.27 Furthermore, heteroduplex molecules may form during the plateau phase of the mixed-template PCR when decreasing primer to template ratios no longer favor primer annealing.33 When heteroduplexes are cloned, the host nick-directed mismatch repair system (MutHLS in Escherichia coli) can convert a heteroduplex into a hybrid sequence by excision repair. As a result of the absence of methylation in the cloned insert, the repair enzymes cannot identify a parent strand and will independently choose either strand as a template for re-synthesis of the complementary base.34 The repaired sequences are therefore composites of the two parent heterologs. This is less likely to occur after round I than round II of a nested PCR and when primers are in excess.26 Because of the low amounts of parasite DNA in many samples, most reactions required a nested PCR to acquire enough DNA to clone, such as for the dbpII. To avoid these PCR artifacts for dbpII, we included only sequences present in two or more samples or sequences that had distinct haplotypes that could not be due to a composite of two other haplotypes from the same sample. Sufficient DNA for cloning msp3 amplicons was generated during the first round of the PCR for some of the isolates. Estimates of the parasite diversity were similar from nest I products compared with that for clones generated from the nest II products (Table 3). Therefore, estimates of genetic diversity for msp3 that used the nest II product may have included some genotypes that were artifacts. This may have overestimated parasite diversity by up to 20% (Table 3), which cannot account for the 2–4-fold greater complexity of P. vivax infection reported here compared with previous studies.
Based on these initial results, we postulate that a large effective P. vivax population is maintained, greater than that for P. falciparum, which reduces inbreeding, sustains continual genetic mixing, and allows recombination to be an important evolutionary force for this parasite to have successfully co-evolved with its human host and persist in the face of intense programs to eradicate it.
Received January 10, 2005. Accepted for publication May 11, 2005.
Acknowledgments: We thank the study participants for their time and P. A. Zimmerman for helpful discussions and suggestions.
Financial support: This study was supported by a grant from the Veterans Affairs Research Service.
* Address correspondence to Jennifer L. Cole-Tobian or Christopher L. King, Center for Global Health and Disease, Case Western Reserve University, Wolstein Research Building, Rm. 4132, 2103 Cornell Rd., Cleveland, OH 44106. E-mail: jlc11{at}cwru.edu, christopher.king{at}case.edu 
Authors’ addresses: Jennifer L. Cole-Tobian, Center for Global Health and Disease, Case Western Reserve University, Wolstein Research Building 4-101, 10900 Euclid Avenue, Cleveland, OH 44106, E-mail: jlc11{at}cwru.edu. Moses Biasor, Papua New Guinea Institute of Medical Research, PO Box 378, Madang 511, Papua New Guinea. Christopher L. King, Center for Global Health and Disease, Case Western Reserve University, Wolstein Research Building 4132, 2103 Cornell Rd., Cleveland, OH 44106 and Research Department, Veteran’s Affairs Medical Center, 10701 East Boulevard, Cleveland, OH 44106, Telephone: 216-368-4817, Fax: 216-368-4825, E-mail: christopher.king{at}case.edu.
Reprint requests: Christopher L. King, Center for Global Health and Disease, Case Western Reserve University, Wolstein Research Building 4-132, 2103 Cornell Rd., Cleveland, OH 44106.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Mideo and T. Day On the evolution of reproductive restraint in malaria Proc R Soc B, May 22, 2008; 275(1639): 1217 - 1224. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Cole-Tobian, P. Michon, E. Dabod, I. Mueller, and C. L. King Dynamics of Asymptomatic Plasmodium vivax Infections and Duffy Binding Protein Polymorphisms in Relation to Parasitemia Levels in Papua New Guinean Children Am J Trop Med Hyg, November 1, 2007; 77(5): 955 - 962. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. COLE-TOBIAN, P. A. ZIMMERMAN, and C. L. KING HIGH-THROUGHPUT IDENTIFICATION OF THE PREDOMINANT MALARIA PARASITE CLONE IN COMPLEX BLOOD STAGE INFECTIONS USING A MULTI-SNP MOLECULAR HAPLOTYPING ASSAY Am J Trop Med Hyg, January 1, 2007; 76(1): 12 - 19. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. BONILLA, L. VALIDUM, R. CUMMINGS, and C. J. PALMER GENETIC DIVERSITY OF PLASMODIUM VIVAX PVCSP AND PVMSP1 IN GUYANA, SOUTH AMERICA Am J Trop Med Hyg, November 1, 2006; 75(5): 830 - 835. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
