| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES |
|---|
. Not being homologous genes, we limited comparisons to within species. P. falciparum (N = 293) had low (He = 0.581) and P. vivax (N = 186) had high (He = 0.845) PLD. A total of 9.5% P. falciparum infections and 26.3% P. vivax infections had COI > 1. Certain allele types were in more mixed infections than expected by chance. The few appearances of new alleles could be explained by stochastic polymerase chain reaction detection or synchronization/sequestration. The results suggest propagation of mixed infections by multiple inocula, not super-infection, implying decade-long opportunity for outcrossing in these mixed infections. |
INTRODUCTION |
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES |
|---|
Several features of malaria transmission at this study site make it particularly interesting. First, the low transmission rates enable us to determine the malaria parasite genetic diversity in discrete (non-overlapping) infections within a relatively newly exposed population.2 Second, the independent characterization of P. falciparum and P. vivax genetic diversity in a sympatric population where there is only one mosquito vector, Anopheles darlingi, allows for more direct interspecies comparisons. 1,2 Finally, malaria infections in this geographic region are surprisingly mild, with > 50% of both P. falciparum and P. vivax infections detected by active case detection being asymptomatic.2 Together, the unique features of malaria in this study region may provide clues to how malaria genetic diversity is maintained in low transmission and the relationship between genetic diversity and the development of immunity and possible drug resistance.
Population-level diversity (PLD) is typically defined by examining genetic variation occurring in variable antigenic loci and/or neutral (purportedly non-antigenic) loci. Because of the low transmission in the Peruvian Amazon, it is also possible to examine PLD occurring in a single mixed-clone infection. Genotyping methods that identify one or more polymorphic, single-copy loci can be used to assess the minimum number of parasite allele-types present within a given infection; this is termed the complexity of infection (COI).3–8 In this study, we studied the PLD and COI for P. falciparum and P. vivax independently, using the P. falciparum merozoite surface protein-1 block 2 region (Pfmsp1-B2) and P. vivax merozoite surface protein-3 (Pvmsp3), respectively.
PfMSP1-B2 is a highly variable repeat-length region near the N terminus of PfMSP1 (PlasmoDB gene ID: PF1475w). Three allelic families (K1, Mad20, and RO33), identifiable by polymerase chain reaction (PCR) of the encoding single copy gene using respective conserved flanking sequences, have been detected globally in all transmission regions.3–13 K1 and Mad20 allelic families can be further distilled by base pair length/size differences observed after electrophoresis because of the expansion and contraction of internal repeating motifs, whereas RO33 is monomorphic in size. 8,9 Both PLD and COI can be defined by the identification of size-specific alleles within each allelic family.
PvMSP3 also contains an internal region comprised of repeating motifs of varying length (PlasmoDB gene ID: PVX_097720). 14,15 Three allelic families have been observed after PCR using P. vivax conserved primers that broadly characterize Pvmsp3: A = 1.8–1.9 kb, B = 1.5 kb, and C = 1.1–1.2 kb. 16–22 Diversity within these allelic families can be further exposed by digesting the PCR product with the restriction enzymes HhaI and AluI.18 Pvmsp3 genotyping patterns can be compiled by using both PCR and restriction fragment length polymorphism (RFLP) to identify infecting alleles and define the PLD and COI.
High PLD (high levels of polymorphism) in both Pfmsp1-B2 and Pvmsp3 has been suspected to arise because of variant-specific immunity, resulting in diversifying selection in the population of parasites. 14,19,21,23–26 Alternatively or additionally, the high PLD could be attributable to asexual rapid diversification of repeat motifs and modest functional constraints on the encoding merozoite surface protein repeat regions. 10,27–31 The high levels of polymorphism in these antigenic loci results in a fine characterization of infecting parasites, which explains their frequent use in malaria genetic diversity studies. Further characterization of the PLD and COI can be achieved by sequencing these regions. Determining PLD and COI of P. falciparum and P. vivax provides insight into the transmission dynamics, population structure, and potential for sexual recombination within the mosquito vector.
In this study, we examined PLD and COI using Pfmsp1-B2 and Pvmsp3. Uniquely, we have weekly sampling to consider the changing parasite genotypes overtime. The low transmission in this Peruvian population suggests that a COI > 1 is caused by one inocula; that is, two genotypes inoculated by one mosquito versus overlapping infections (super-infection). The purpose of this study is to establish the baseline genetic diversity levels in this region of the Peruvian Amazon. We hypothesized that there would be low P. falciparum and low P. vivax PLD and COI. We propose a molecular test and discuss the evolutionary consequence of mixed infections without super-infection.
|
MATERIALS AND METHODS |
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES |
|---|
In the active case detection, blood slides from individuals who had a body temperature of 
38.3°C, reported having a high fever within 2 days, or had hematocrit < 30% pcv, had their blood slides read by expert microscopists within 1 day and were treated within 1 day if positive. If an individual was asymptomatic, there could be 6 days before reading the microscopy slide. At the next scheduled visit (1 week later), another blood sample was collected, and any individual who was found with malaria parasites on the week before visit had their blood slide read immediately. Therefore, asymptomatic individuals had more than one blood slide collected during a given infection before treatment. Treatment regimens are reported in Branch and others.2
DNA extraction and testing for Plasmodium species. All blood samples, positive or negative by microscopy, underwent DNA extraction by Qiagen DNeasy Blood and Tissue Kits and were tested for presence of Plasmodium species using a nested-multiplex PCR method targeting DNA encoding the small subunit ribosomal protein (ssrDNA) (Qiagen Inc., Valencia, CA). 32 For this species-specific ssrDNA PCR diagnosis and PCR for genotyping (described below), PCR mastermixes were comprised of primers at the indicated concentration (Integrated DNA Technologies, Coralville, IA): 2.0 mmol/L concentration of deoxynucleoside triphosphate (dNTP) mixture (Invitrogen, Carlsbad, CA); MgCl2 at a concentration of 1 mmol/L (Promega, Madison, WI); 5x PCR buffer, Taq polymerase (GoTaq Flexi; Promega), molecular grade water, and blood extracted DNA (genomic DNA adjusted to 10–20 ng/µL) or externally amplified (PCR amplicon) DNA. All PCRs were conducted in an Eppendorf Mastercycler ep (Westbury, NY).
PCR genotyping Pfmsp1-B2. Pfmsp1-B2 genotyping was attempted on all available samples that were microscopy and/or PCR positive for P. falciparum (293 P. falciparum infection-months). Previously reported genotyping methods were used for main allelic family analysis, using all permutations of primers to detect any allele families that were one allele type in the 5' direction and a different allele type in the 3' direction. 8,10
PCR products for the K1, Mad20, and R033 allelic families were observed; we did not observe the Mad20-R033 recombinant (MRrec) allele. 10 The amplicons from each PCR reaction were individually visualized by electrophoresis using a 1.5% concentration of UltraPure Agarose 1000 (Invitrogen). The alleles were characterized by allelic family and size of the amplified product using the Gel Doc gel analysis system (Bio-Rad Laboratories, Hercules, CA). The lengths were assigned by rounding up or down to the nearest 10 bp characterization based on the 25 bp ladder (Invitrogen, Carlsbad, CA). Final allele assignment was determined by performing electrophoresis a second time, arranging the samples in order of base pair length determined from the first electrophoresis and confirmed by sequencing. Samples with band sizes between 250 and 300 bp for both K1 and Mad20 allelic families (not shown) were found to be artifacts of the first external PCR and therefore were removed from the analysis.
PCR genotyping Pvmsp3.
The nested-PCR protocol of Bruce and others 16 was used on samples P. vivax positive by microscopy and/or PCR. Primers were used at a concentration of 5 µmol/L. For the further-discerning RFLP assay, 8 µL of the PCR products was digested with HhaI and AluI enzymes, individually, overnight at 37°C (4 units of enzyme/reaction; Promega). The PCR-amplified products underwent electrophoresis on a 1.8% agarose gel (Low Range Ultra Agarose; Bio-Rad). Fragment sizes were determined using the Gel Doc gel analysis system (Bio-Rad). Additionally, gel images were evaluated by hand and compared with the software-based analysis to ensure fragment calling consistency. The resulting fragment sizes were assembled into an Excel file where there was a column for all possible sizes (one bin for every 10 bp). Each RFLP pattern was represented by having fragment sizes present/absent in the array of bins. Patterns with fragment sizes that overlapped or were within 20 bp across fragments were considered to be the same pattern.
Sequencing of Pfmsp1-B2 and Pvmsp3.
We sequenced a selection of infections (selecting without any apparent bias) representing the alleles detected for both P. falciparum and P. vivax. Using the product of the second PCR reaction (second in the two-step nested method), the amplified PCR products were isolated (Gel Extraction Kits; Qiagen). In all cases, the 5' and 3' nested-PCR primers were used in the sequencing reaction at a concentration of 2 µmol/L. Sequencing was performed on an ABI 3730 Genetic Analyzer (Applied Biosystems, Foster, CA).
Analysis of PLD and COI.
We defined infections as one or more malaria-positive sample point(s) during the follow-up, which in many cases included time periods of the weekly active case detection. We considered all sampled time points within 1 month before and after the infection. The sampling could begin before the initial malaria positive sample or after, with an overall sampling time frame of 
30 days (4 weeks of sampling follow-up planned). For this reason, infections are called "infection-months," and this is the denominator used when considering number of infections followed and frequency of the different parasite alleles detected.
The expected heterozygosity, a measurement PLD, was calculated by the following formula: He = 1/(1 – N)(1 – 
p2i), with p equal to allele frequency and N equal to the number of alleles detected. The He was used to quantify the overall genetic diversity in each species using the respective genotyping targets and also to compare the genetic diversity within each allelic family within each species. A COI > 1 could be determined by observing multiple alleles within a single sample or by concatenating alleles observed in an infection-month.
We considered the transmission level by using both the active and passive case detection prevalence determined for the whole Zungarococha community in an earlier study2 and by calculating the maximum transmission rate from the active and passive case detection prevalence in these 1,735 individuals (dividing infections detected in these individuals by 1,735, with the actual rate being calculated by determining the rate of becoming infected during the active sampling). The range of these measures was 0.022–0.069 infections/person/month for P. falciparum and 0.065–0.107 infections/person/month for P. vivax. The highest transmission level was when considering the active case detection transmission rate calculation, which was 0.069 and 0.107 infections/person/month for P. falciparum and P. vivax, respectively. Therefore, to test the probability of mixed infections, we used this high-end transmission rate calculation.
|
RESULTS |
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES |
|---|
There were 308 P. falciparum infection-months identified within 2003–2004, of which 293 were selected for genotyping in this study. Of these 293 infection-months selected, 180 were identified by microscopy through active case detection, 63 were sub-microscopic and identified by 18 ssrDNA methods through active case detection, and an additional 50 infection-months were identified by microscopy and/or 18 ssrDNA methods through passive case detection (passive case detected infection-months do not have follow-up data; Table 1
).
|
). Of those infection-months that had multiple sample points, new alleles were detected in the successive sample points in only 2 of the 243 P. falciparum (0.8%) and 17 of the 136 P. vivax (12.5%) infection-months where there were multiple sample points. Most P. falciparum and P. vivax infection-months were single allele infections and/or did not have new alleles detected over time; as will be described in detail below, this could be explained by stochastic PCR detection, allele synchronization/sequestration, or, in the case of P. vivax, relapse.
Pfmsp1-B2, population diversity.
The K1, Mad20, and RO33 allele families that have been detected globally in prior studies were detected in Zungarococha. We observed two K1 allele types (170 and 195 bp), two Mad20 allele types (200 and 210 bp), and one R033 allele type (140 bp). There were a total of 322 parasite allele detections: 202 (62.7%) K1, 112 (34.8%) Mad20, and 8 (2.5%) RO33. The most frequent K1 allele type detected in this population was K1-195 (N = 176; 54.7%); K1-170 was observed less frequently (N = 26; 8.1%). Of the Mad20 allelic family, the Mad20-210 allele type was most frequent (N = 87, 27.0%), whereas Mad20-200 was observed less frequently (N = 25, 7.8%; Figure 1).
|
To consider accuracy/resolution of the genotyping results, we sequenced 11 K1-170 (GenBank accession numbers: FJ612009–FJ612019), 18 K1-195 (FJ612020–FJ612037), 2 Mad20-200 (FJ612038 and FJ612039), 24 Mad20-210 (FJ612040–FJ612063), and 2 RO33-140 (FJ612064 and FJ612065). The sequences coincided with the genotyping results; there was only one polymorphism, a synonymous nucleotide substitution within a Mad20-210 variant at position 101 (FJ612041).
Pfmsp1-B2, complexity of infection.
Of 293 genotyped P. falciparum infection-months, there were 265 (90.5%) single genotype infections and 28 (9.5%) mixed genotype infections (Table 2). Of the mixed infections, 4 (1.4%) were mixed alleles of the same allelic family and 24 (8.2%) were mixed alleles of at least two allelic families. The COI was equal to 1 in 90.5% of infection-months, COI = 2 in 9.2% of infection-months, and COI = 3 in 0.3% of infection months.
|
). The Mad20-200 allele type was observed 24 times in this population; of those, it was observed mixed with the K1-195 allele-type in seven instances. The expected rate of a K1-195 + Mad20-200 mixed infection occurring by super-infection would be 0.0018 (or [(0.5222 x 0.0512) x 0.069]), respectively. The observed rate of these two alleles occurring concurrently in the same infection-month was 0.0239, which was significantly greater than the expected rate of two separate inoculating events (P < 0.01, Fisher exact test). Similarly, of the eight times R033 was observed in this population, six times was in a mixed infection with K1-195. The expected rate for a K1-195 + RO33 mixed infection occurring by super-infection would be 0.0001 (or [(0.5222 x 0.0034) x 0.069]); yet, this combination was observed in 0.0204 of the infections, which was significantly greater than expected (P < 0.01, Fisher exact test). Overall, inter-family mixed infections (mixed alleles of different allelic families) were observed 8.64 times more frequently than intrafamily mixed infection (mixed alleles of the same allelic family).
Pvmsp3 population diversity.
Initially, we identified allelic families by size differences in the undigested PCR product. Further characterization was made by RFLP digestion of these amplified products with HhaI and AluI digestive enzymes. Genotyping results (alleletypes) were described by a combination of both PCR and RFLP pattern codes: "Allelic Family_Hha I pattern_Alu I pattern" (Figure 2A and 2B). By this method we observed all three Pvmsp3 allelic families found globally (A = 1.9 kb, B = 1.5 kb, C = 1.2 kb). There were 198 of 222 allele detections (from 168 of 186 infection-months) where the PCR-RFLP combination pattern (Allelic Family_Hha I pattern_Alu I pattern) suggested a single, discernible allele present. These 198 alleles were used when considering the allele frequency (Table 3). We did not include patterns suggesting multiallele detections (the 24 allele detections shown in bold in Table 3) when we could not discern the components of mixed infections.
|
|
allele frequency calculation (excluding 18 individuals who had a RFLP pattern that could not be discerned to a single allele type), there were 172 (86.9%) A-type, 20 (10.1%) B-type, and 6 (3.0%) C-type alleles found (number exceeds total because there were identifiable mixed infections; Figure 3). In this analysis, significant intrafamily diversity was found within the A-type allelic family (23 alleles), whereas little diversity was observed in the B- and C-types (1 and 3, respectively).
|
To determine how the 44 PCR-RFLP genotype combinations reflected sequence differences, we sequenced 68 samples representing 14 of the 44 different allele types. Of the 68 samples, there were 40 different sequences observed (FJ612066–FJ612105), where the presence of at least one non-synonymous substitution defined a sequence as "unique." Sequences were correlated to the original PCR-RFLP allele types observed by the selection of eight regions of variability (Figure 4 Supplementary Data Figure 1 available online at www.ajtmh.org). Of the 40 sequences, there were eight distinct groups. These eight sequence groups were well described by the PCR-RFLP study (six sequences could not be included because of incomplete sequencing data). If a sequence within a group defined a unique allele type or if a genotype defined a sequence, this was considered concordant. The PCR-RFLP genotype combination and sequencing data was 84.9% concordant (Figure 4; Supplementary Data Figure 1), indicating that > 80% of the variation within the sequence is reflected in our PCR-RFLP pattern. Although sequencing methods provide a more accurate assessment of the PLD present in Pvmsp3, this PCR-RFLP genotyping method allows for the differentiation between single genotype and mixed genotype infections, which cannot be resolved by sequencing methods alone.
|
complexity of infection.
In total, there were 137 (73.7%) single genotype infections and 49 (26.3%) mixed genotype infections (Table 3). Of the mixed infections detected, 11 (5.9%) were due to the detection of more than one allelic family (determinant 1), 23 (12.4%) were due to alternating alleles concatenated into the same infection-month (determinant 2), and 15 (8.07%) were mixed as a result of the summation of RFLP fragments exceeding the allelic family size by 2 SD (determinant 3). A COI = 1 was observed in 73.7%, a COI = 2 in 15.6%, and a COI = 3 in 2.7% of the infections (frequency excludes infections where allele types could not be surmised). There were some alleles detected together in mixed infections that were observed in this population more than expected by super-infection. Of the 31 times that A_5_1 and the 32 times that A_6_4 were detected in this population, there were 12 infections where they were observed mixed. The expected rate of A_5_1 + A_6_4 occurring by super-infection (where allele frequency is a factor of the transmission rate ([0.1667 x 0.1720] x 0.107) was 0.0030. The observed rate of A_5_1 + A_6_4 mixed infections occurring in this population was 0.0645, which is greater than expected by super-infection (P < 0.0002, Fisher exact test). Intrafamily mixed infection-months were observed 5.10 times more frequently than inter-family mixed infection-months.
New alleles detected over time.
There were 58 P. falciparum infection-months that could be evaluated based on more than one genotype result taken over multiple sampling points (Table 1). Forty-eight (82.8%) of these had a COI = 1, 8 (13.8%) has a COI > 1 with no new allele types detected over time, and 2 (3.5%) had a COI > 1 with new allele types detected over time (Table 4, left). Of the 40 P. vivax infections we could evaluate over time, there were 19 (47.5%) that had a COI = 1 (Allelic family_Hha I PCR-RFLP patterns), 4 (21.1%) with a COI > 1 with no new allele types detected over time, and 17 (42.5%) with new allele types detected over time (Table 4, right).
|
i[p(1 – p)x x], with p equal to allele frequency of the ith allele and x equal to the maximum transmission rate observed in this study (P. falciparum = 0.069; P. vivax = 0.107). The expected frequency of super-infection for P. falciparum and P. vivax was 4.24% and 9.01%, respectively. Testing the hypothesis of no super-infection (0), neither the P. falciparum nor the P. vivax expected frequency of super-infection was significantly different from zero (P = 0.24 and P = 0.12, respectively, Fisher exact test). The possibility of relapsing infections in P. vivax does not limit our conclusion that, over the infection-month, the appearance of new alleles by super-infection is expected to be rare. |
DISCUSSION |
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES |
|---|
, respectively. This genetic diversity study was not meant to compare PLD between infecting species as in Ord and others, 33 which used homologous apical membrane antigen-1 (AMA1) encoding genes to elegantly characterize and compare the PLD between species in the mesoendemic Venezuelan Amazon. Although Pfmsp1-B2 and Pvmsp3 are not homologous encoding genes, we used these markers because of their frequent use in earlier studies and their shown value as highly polymorphic markers that make fine distinctions between different parasites to define allele types. By defining the PLD and COI of P. falciparum and P. vivax species, correlative studies can be performed by drawing on this reported diversity as a baseline; for example, it provided the foundation for the future characterization of PLD and COI using neutral (purportedly non-antigenic) loci markers. Using Pfmsp1-B2 as a marker for genetic diversity in P. falciparum, we quantified PLD and found low levels of genetic diversity within the entire population (He = 0.581) and within each allelic family (K1 = 0.221, Mad20 = 0.283, and RO33 = 0.000). The Pfmsp1-B2 diversity was significantly less in this Peruvian cohort than has been observed in hyper/holoendemic geographic regions, such as Senegal,7 Papua New Guinea,3 and western Kenya.8 For example, in western Kenya, during a time with > 300 infected P. falciparum bites/person/night, there were as many as 21 P. falciparum allele types defined by Pfmsp1-B2 at a population-level and up to 6 Pfmsp1-B2 allele types detected in one infection.8 Other studies in South America have reported similar or even lower PLD than observed in this study. 9,12–14,34 Recently, increased Pfmsp1-B2 diversity was reported in a collaborative study using some samples from our cohort. 12 However, the results from this collaboration are not directly comparable because the previous study 1) included samples from different communities and years and 2) only reported genetic diversity based on sequencing methods. Defining PLD solely by sequencing eliminates the possibility of detecting mixed genotype infections and may result in the exclusion of rare alleles that are found paired in the population, such as RO33-140 paired with K1-195. In our study, sequencing was used to further characterize the genetic diversity on a codon level and not as the primary method of allele detection.
We found high overall P. vivax PLD, quantified by expected heterozygosity (He = 0.845) and similarly high levels of intra-family PLD (A-type = 0.808, C-type = 0.600) with the exception of the B-type allelic family (He = 0.000). Our results were consistent with other Pvmsp3 PLD studies. 16–22 Relatively high PLD is expected within this polymorphic marker even in hypoendemic regions; from non-synonymous amino acid substitutions to the expansion and contraction of the sequence by inclusion or exclusion of minor indels, increased diversity in Pvmsp3 is a consequence of balancing selection or slipped-strand mismatch repair during replication. 21 In hyperendemic regions like Papua New Guinea (PNG), Bruce and others 16 reported that there were as many as 11 different alleles types detected in only 12 infections (133 bites/person/year) by a PCR-RFLP pattern defining method. More recently, Ord and others reported nine different Pvmsp3 alleles from 131 infections in the mesoendemic Venezuelan Amazon using sequencing methods. In our study, we found 44 different allele types from 186 infection-months defined by the PCR-RFLP method and were even able to identify many complex infections. Sequencing confirmed the high P. vivax PLD and also showed additional genetic variation, permitting the identification of sub-alleles. Based on sequence results, we were able to cluster the 34 strains into eight different groups based on relatedness. Ord and others 21 suggested that A-type alleles are the putative progenitors of the smaller B- and C-type alleles. Based on the correlative PCR-RFLP/sequencing methods in this study, we clearly observed relatedness between these allelic families (Figure 3) and submit that perhaps increased selection on A-type alleles decreases selection on the B- and C-type alleles, explaining their lack of diversity.
The few detections of new genotypes over the month sampling in this study might be attributable to 1) a new infection (super-infection), 2) relapsing infection, or 3) PCR not detecting all alleles that were inoculated in one mosquito bite (stochastic PCR results or alternating alleles caused by parasite densities in the host at the time of sample collection). 18,35,36 The increased frequency of new alleles detected over time in P. vivax might be attributable to infection by relapse, despite treatment with Primaquine. However, because of 1) the low transmission rate (P. falciparum = 0.13 and P. vivax = 0.39 infections/person/malaria season), 2) high frequency of single genotype infections (P. falciparum = 90.5%, P. vivax = 73.7%), 3) the observation of allelic pairs appearing together more frequently that expected by chance, and 4) the few observations of new alleles introduced over time (P. falciparum = 2/58, P. vivax = 17/40), we predict that mixed genotype infections are likely the result of a multiple inoculation event. We calculated the expected rate of super-infection causing mixed genotype infections in this population based on the observed allele type frequencies and transmission rate; this calculation was even biased toward the null hypothesis by using the maximum observed transmission rate in this study. We found that the mixed genotype infections in this study are likely propagated by a single multiple inoculation event rather than by super-infection. Further study of the propagation of mixed genotype infections by a single inoculation event is necessary to confirm the intermittent role of super-infection in this region. Our future studies will investigate the homogenization of genomes as a consequence of alleles existing together over time and the role of natural selection on the alleles within these mixed infections.
Received March 13, 2009. Accepted for publication July 4, 2009.
Acknowledgments: The authors thank the Zungarococha community members and authorities for their ongoing commitment to the Malaria Immunology and Genetics in the Amazon (MIGIA) Study. The authors thank the Director of the Universidad Cayetano Heredia, Instituto de Medicina Tropical Alexander von Humboldt, Dr. Eduardo Gotuzzo, as well as Dr. Alejandro Llanos and Dr. Humberto Guerra, for their coordination and guidance. The authors thank the Universidad Nacional de la Amazonía Peruana, Laboratorio Naturales Antiparasitarios de la Amazonía, Dr. Lastenia Ruiz, and investigators for their coordination and technical assistance in completing a portion of the Pfmsp1-B2 genotyping. The authors thank Dr. Willy Diaz for patient care and follow-up, Freddy Alava for enrollment and follow-up, Ever Alvarez and Anibal Sanchez for microscopy, Zoila Reategui and Elva Sanchez for patient visits, Dania Vela and Noelia Escobar for preparation of materials, separation and storage of blood samples, and data entry, Dr. Julian Rayner for consultation and manuscript review, and Dr. Robert Oster for his review of the statistics reported in this manuscript.
Financial support: This study was supported by RO1 Grant AI064831 from the National Institute of Health/National Institute of Allergy and Infectious Disease (2005–present) and pilot funding from the Gorgas Memorial Institute (2003) and Sparkman International Center for Public Health (2003).
* Address correspondence to OraLee H. Branch, Department of Medical Parasitology, New York University, 341 E. 25th Street, OPH, New York, NY 10010. E-mail: OraLee.Branch{at}nyumc.org 
Note: Supplementary Data Figure 1 can be found online at www.ajtmh.org.
Authors’ addresses: Patrick L. Sutton and OraLee H. Branch, Department of Infectious Diseases, University of Alabama at Birmingham, Birmingham, AL 35294. Victor Neyra, Instituto de Medicina Tropical "Alexander von Humboldt," Universidad Peruana Cayetano Heredia, Lima, Peru. Jean N. Hernandez, Universidad Nacional de la Amazonia Peruana–Laboratorio de Investigaciones Productos Naturales Anti-parasitarios, Iquitos, Peru. OraLee H. Branch, Department of Medical Parasitology, New York University School of Medicine, New York, NY 10010.
|
REFERENCES |
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES |
|---|
locus of Plasmodium vivax: global and local diversity. Am J Trop Med Hyg 61: 518–525.[Abstract] is limited to specific domains. Parasitology 125: 393–405.[Web of Science][Medline] is a reliable marker for population genetic analysis of Plasmodium vivax. Malar J 5: 53.[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
