Background

Genetic crosses between Plasmodium falciparum lines with distinct phenotypes have supported the identification of genes involved in malaria drug resistance, immune evasion, host infection, and disease pathogenesis. However, reproducing the malaria parasite life cycle in the laboratory is complex and expensive [1]. The process requires: production of infectious gametocytes from parasite cultures; infection of Anopheles mosquitoes with these gametocytes for cross-fertilization and generation of recombinant sporozoites (Plasmodium are haploid except for a brief diploid phase from zygote to oocyst); inoculation of the sporozoites into a vertebrate host for production of liver stage parasites; and recovery of blood-stage recombinant progeny (segregants) that emerge from the liver and propagate in erythrocytes. Five P. falciparum genetic crosses have been completed over the past 40 years. Four of these crosses were performed through chimpanzees (Table 1) [2,3,4] (and unpublished work). More recently, a fifth cross was produced through immune-deficient mice engrafted with liver and erythrocytes from humans [5].

Table 1 Four Plasmodium falciparum genetic crosses in nonhuman primates
Full size table

Recombinant progeny from the four non-human primate (NHP) crosses have led to fundamental research advances in diverse studies of malaria phenotypes (Additional file 1). Protecting the integrity of the P. falciparum parent and progeny lines has been critical to their value as a research resource in support of these and other advances. In a laboratory setting where multiple culture flasks are handled, investigators must continually guard against the risks of cross-contamination or mislabeling [6, 7]. Vigilance is imperative, and it is necessary to perform regular checks of parasites recovered from cryopreserved vials, in long-term cultures, and before cryopreservation for legacy stocks.

Progeny of the P. falciparum genetic crosses have been typed by various methods. HB3×3D7 progeny were originally typed by restriction fragment length polymorphism (RFLP) methods in which parasite DNA cut by restriction enzymes is separated on agarose gels, blotted to a membrane, and probed for DNA fragment size differences [8, 9]. HB3×Dd2 progeny were typed by RFLP [10] and microsatellite analysis [11, 12], a method that utilizes polymerase chain reaction (PCR) to amplify stretches of short polymorphic repeats that can be highly variable between parasite lines. Microsatellite and tandem repeat regions that differ by many base pairs can be visualized on agarose gels, whereas microsatellite regions that differ slightly in size are more appropriately analyzed by polyacrylamide sequencing gel or capillary electrophoresis. 7G8×GB4 progeny were typed using microsatellites and high-density oligonucleotide tiling arrays, which differentially hybridize with fragments of DNA containing single nucleotide polymorphisms (SNPs) [13, 14]. 803×GB4 progeny have been typed using microsatellites and a DNA microarray containing 17,000 potential SNPs (unpublished). More recently, progeny from the HB3×3D7, HB3×Dd2, and 7G8×GB4 crosses were analyzed by full genome sequencing [15].

A typing method must be fast, reliable, and inexpensive for regular checks of parasite cultures in the laboratory. RFLP methods require large amounts of DNA, and probe preparation and blotting after gel electrophoresis is time consuming. DNA microarrays also require large amounts of DNA and depend on skilled processing and analysis with specialized equipment. Modern genome sequencing can give snapshots of total variation, but cost, turnaround time, and analysis remain prohibitive for routine typing checks. In contrast to these methods, microsatellite analysis has a production time of less than 1 day, requires only small amounts of DNA, and is relatively inexpensive [12].

A set of 12 microsatellite markers that distinguishes progeny of the four NHP P. falciparum crosses is presented here. Use of these markers can be extended to diverse parasite lines and help to maintain their integrity and identity among different laboratories.

Methods

Parasite cultivation and genomic DNA

Erythrocytes obtained weekly from Interstate Blood Bank (Memphis, TN) or Virginia Blood Services (Richmond, VA) were washed with filtered RPMI 1640 media (KD Medical, Columbia, MD) and stored at 50% hematocrit in a 4 °C refrigerator until use. Complete culture media (cRPMI) was comprised of 1% Albumax II (Life Technologies, Carlsbad, CA), 0.21% sodium bicarbonate (KD Medical, Columbia, MD), and 20 µg/mL gentamicin (KD Medical, Columbia, MD) in RPMI-1640, which contains 25 mM HEPES and 50 µg/mL hypoxanthine. Parasites were cultivated at 5% haematocrit between 0.5–3.0% parasitaemia at 37 °C under an atmosphere of 90% N2, 5% CO2, and 5% O2. When cultures reached approximately 4% parasitaemia with mostly mature stages, 10 mL volumes were pelleted at 2500 rpm, and the parasitized cells were treated with 0.15% saponin (Amresco, Solon, Ohio) in 1× phosphate buffer solution (PBS) (90 g/L sodium chloride, 1.44 g KH2PO4/L, 7.95 g NA2HPO4/L) (KD Medical, Columbia, MD) for 3 min at room temperature. Parasites were recovered from the resulting host cell lysate by centrifugation at 4000 rpm, washed three times with 1× PBS, and stored at − 20 °C. Some samples of genomic DNA from the 803×GB4 progeny were obtained from smaller culture volumes (~ 2 mL). Genomic DNA was extracted from thawed samples by the phenol:chloroform method [16] or by the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). DNA concentration was determined using a NanoDrop ND-100 Spectrophotometer.

Polymerase chain reaction

PCR was performed in 96-well plates in an Applied Biosystems SimpliAmp Thermal Cycler with the following conditions: denaturation for 2 min at 94 °C; 42 cycles through 94 °C for 20 s, 45 °C for 10 s, 42 °C for 10 s, and 60 °C for 30 s; and a final extension at 60 °C for 5 min. Reaction volumes of 25 µL included: 1 µL of template DNA at a concentration of approximately 20 ng/µL, 0.5 µL of each 10 µM forward primer and 10 µM reverse primer with a 5′ 6-FAM fluorescence modification (Eurofins, Louisville, KY), 12.5 µL MyTaq Mix (Bioline, Taunton, MA), and 10.5 µL PCR qualified water (Quality Biological, Inc., Gaithersburg, MD).

Microsatellite analysis

Amplified DNA was diluted 100× in an optical 96-well plate with 10 µL of Hi-Di Formamide (Life Technologies Corporation, Carlsbad, CA) and 0.2 µL GeneScan 500XL ROX Size Standards (Life Technologies Corporation, Carlsbad, CA) per well. Plates were processed in a Hitachi 3730xl DNA Analyzer and evaluated using GeneMapper version 4.1 (Applied Biosystems, Foster City, CA). Base pair sizes of microsatellite alleles were recorded at the peak heights of their signals presented on electropherograms.

Mixed DNA samples experiment

Stock 803 and GB4 genomic DNA dilutions were combined in 1.5 mL Eppendorf tubes (Denville Scientific Incorporated, Holliston, MA) in 803:GB4 DNA ratios of 5:95, 10:90, 20:80, 50:50, 80:20, 90:10, and 95:5. These mixed DNA samples were typed as described above using markers TAA87, TA127, 1451458, and C3M69.

Results and discussion

Thirty-two candidate microsatellite markers were identified from a large collection originally developed to differentiate HB3×Dd2 and 7G8×GB4 progeny (Table 2 and Additional file 2) [4, 11]. Selection criteria included: (1) physical location in the genome to avoid linkage disequilibrium and cover the most chromosomes; and (2) PCR product size that discriminated between the parents of each cross by more than 3 base pairs. Of these 32 microsatellite markers, 14 distinguished the 3D7, Dd2, HB3, 7G8, GB4, and 803 parental lines from one another. To facilitate efficient PCR in the 12-column format of 96-well plates, a final set of 12 markers was selected for typing. Table 2 lists the primer sequences, genome locations, and expected PCR product sizes of the 12 microsatellites.

Table 2 A set of 12 microsatellite markers that distinguishes parents and recombinant progeny of four Plasmodium falciparum crosses
Full size table

The 12-marker typing set identified distinguishing patterns of inheritance (fingerprints) in segregant collections of 15 progeny clones from HB3×3D7, 32 from HB3×Dd2, 33 from 7G8×GB4, and 81 from the new 803×GB4 (Additional file 3). Since peaks on the DNA analyzer can shift slightly depending on machine running conditions, parental control DNAs were included in all typing determinations along with the size standards.

Minor mixed or contaminated parasite subpopulations can be missed by microsatellite typing when they represent < 20% of the overall parasite population [17, 18]. In further experiments with microsatellite markers to detect cross-contamination, known amounts of 803 and GB4 genomic DNA were mixed and tested with markers TAA87, TA127, 1451458, and C3M69, all of which have alleles that differ by 10 base pairs or more. The microsatellite signals from minor quantities of DNA differed greatly in level and, in some cases, were difficult to detect in the individual scans (Fig. 1). Peak height ratios were consistent with greater amplification of smaller alleles of the four individual markers, in agreement with previous findings of a general bias to smaller amplicons [18] (see representative TAA87 and TA127 data in Additional file 4).

Fig. 1
figure 1

Typing results from mixtures of 803 and GB4 DNA tested with microsatellites TAA87, TA127, 1451458, and C3M69. The 803 DNA has a longer repeat sequence than GB4 in microsatellites TAA87 and TA127, whereas GB4 has a longer microsatellite sequences than 803 in microsatellites 1451458 and C3M69. Solid and dashed lines represent the 1/3 and 1/4 cut off levels, respectively

Full size image

While microsatellite analysis is an efficient method to check parasite line integrity, our results show that minor populations may still be missed. Filters that disregard minor peaks less than 1/4 or 1/3 the size of the major peak are sometimes used in population genetic studies, so the use of these filters in microsatellite analysis may give false impression of clonality with mixed parasite populations [19,20,21,22]. Unfiltered searches for minor peaks with multiple markers are therefore recommended for detection of allelic variations indicative of mixed or contaminated parasite subpopulations.

The polymorphism of these microsatellite markers can be useful for genetic population studies [19, 20, 23, 24] as well as future typing applications in genetic crosses. Although the microsatellite markers were stable in this study, infrequent de novo size variations are possible; an example has been reported in a clonal parasite population after long-term culture [11].

Comparison of sample typing patterns to published fingerprints presently relies on human ability to recognize visual patterns and find differences, and potential for error remains in assessment of clonality or disparity between lines. Automated searches may help to improve these comparisons in the future, perhaps with algorithms of pattern matching against microsatellite databases.

Conclusions

A set of 12 microsatellite markers is presented to distinguish genetically related P. falciparum progeny of laboratory crosses, with useful fingerprints to maintain the integrity of diverse parasite lines in and between research laboratories.