Development of a TaqMan Allelic Discrimination Assay for detection of Single Nucleotides Polymorphisms associated with anti-malarial drug resistance

Samples used in this study were obtained from a Phase IIb paediatric clinical trial conducted between March 2005 and April 2006 at the KEMRI/Walter Reed Project, Kombewa Clinic in the Kombewa Division of Kisumu District, Nyanza Province, Western Kenya. The study was approved by the Ethical Review Committee of the Kenya Medical Research Institute, Nairobi, Kenya and Walter Reed Army Institute of Research (WRAIR) Institute Review Board, Maryland, USA. The details of this study have been described elsewhere [26]. Genomic DNA from the clinical sample was extracted from whole blood using the QIAamp DNA Blood Mini Kit (Qiagen, Valenica CA, USA) as recommended by the manufacturer.

Genomic DNA from three P. falciparum laboratory strains was used in this study. Genomic DNA from 3D7 and 7G8 strains was extracted from cultures grown at WRAIR using QIAamp DNA Blood Mini Kit. K1strain genomic DNA was obtained from Malaria Reagent Repository Resource http://​mr4.​org, MR4.

Thirty SNP sequences (~100 nucleotides on the left frank and ~100 nucleotides on the right frank of the allele of interest) associated with anti-malarial drug resistance were submitted to Applied Biosystems (AB) website for analysis of their suitability to be used in SNP assays. After submitting sequences to the custom TaqMan assay design tool on the AB website, twenty five sequences returned SNP assays that were considered to be good candidate designs. The following SNP assays were selected for further analysis: pfdhfr codons A16V, S22N, N51I, C59R, S108N, and I164L; pfmdr1 codons N86Y, Y184F, S1034C and N1042D; and pfdhps codon A581G. TaqMan-MGB genotyping assay mix for each of these assays was ordered from AB. Each of these TaqMan-MGB genotyping assay mix contained a forward and reverse primer, one probe that perfectly matched to the wild-type sequence variant labeled with VIC and the second probe matched to the mutant (SNP) sequence variant labeled with 6-carboxyfluorescein (FAM) (Table 1). Probes contained a non-fluorescent quencher and a minor groove binder moiety (MGB) which allows for shorter probe sequences to be designed. TaqMan-MGB genotyping assay mixes were supplied at 40X concentration.

A working master mix was prepared that contained 0.25 μL of TaqMan-MGB genotyping assay mix (20X), 2.5 μL TaqMan Genotyping Master Mix (Applied Biosystems, Carlsbad CA, USA) and 2.25 μL of water or 1.25 μL water and 1 μL of human genomic DNA at 50 ng. Human genomic DNA was added in reactions which were ran with plasmid DNAs as the template. Each reaction contained 5 μL of working master mix and up to 2 μL of template DNA. Template DNA used in these experiments was either plasmid DNA or genomic DNA. In some experiments, two different plasmid DNAs were used. In such instances, 0.5 μL or 1 μL of each plasmid DNA was added to the 5 μL of the working master mix bringing the final volume of the reaction to 6 or 7 μL. Extensive experiments (to be reported elsewhere) have been done which show that such differences in volume and concentration of real-time PCR reactions do not alter assay performance. Plates were loaded into Applied Biosystems 7500 Fast Real-Time PCR System and AB 7500 v 2.0.5 software was used to run the genotyping assay experiment following the default standard Allelic Discrimination genotyping assay protocol. The initial step of this protocol includes pre-reading of the plate where the background florescence is recorded followed by AB standard PCR protocol of 95°C for 10 min, 95°C for 15 sec and 60°C for 1 min, repeating steps 2-3 for 40 cycles. The post-read step follows after completion of the PCR step where a post-read is performed. The software analysed the before and after florescence level and calculates normalized dye fluorescence (ΔRn) as a function of cycle number for Allele1 (wild-type) or Allele2 (mutant). Based on this number (which is a relative number of the possible outcomes), the software makes an automatic call of either Allele1 (homozygous1/1), Allele2 (homozygous2/2) or heterozygous (1/2). In some instances however, the software does not make automatic calls and returns a call of undetermined which can be due to low or no change of fluorescence. Also, the system requires large enough sample number in order to make automatic calls (from communicating with AB technical help line, ~fifteen or more samples are required for the software to make automatic call). However, in some instances there is no clear explanation why the software does not make automatic calls but with proper controls in place, calls can be made manually by assessing the calculated normalized dye fluorescence (ΔRn) as a function of cycle number and the cycle threshold (CT) values.

Primers were designed (using Primer Express 3.0 software) that amplified amplicons of interest (Table 2). DNA was amplified from genomic DNA of either P. falciparum laboratory strain samples or clinical samples. The amplification reaction mixture contained 1 × PCR gold buffer with MgCl₂ in a final concentration of 4 mM, 0.2 mM deoxynucleoside triphosphates, and 0.4 μM of each primer. Reactions were carried out in 20 μL reactions containing 1 μL genomic DNA and 1 U AmpliTaq Gold DNA Polymerase with GeneAmp (Applied Biosystems, Carlsbad CA, USA). Cycling conditions were 95°C for 10 min followed by 40 cycles of 95°C for 15s, 52°C for 10s, and 72°C for 30s. Amplification of the PCR fragments was confirmed by running part of the PCR reaction on a 2% agarose gel and staining with ethidium bromide. PCR fragments were purified using QIAquick PCR purification kit (Qiagen, Valenica CA, USA) following manufacturer's instructions. Purified PCR fragments were cloned into pCR 2.1-TOPO vector using TOPO TA cloning kit (Invitrogen, USA) following manufacturer's recommendations. One Shot TOP10 Chemically or Electrocomp competent cells (Invitrogen, USA) where transformed with cloned vector following manufacturer's recommendations. These reactions were plated out on kanamycin plates and single colonies were grown in cultures overnight. Plasmid DNAs were purified from overnight cultures using QIAgen spin miniprep kit (Qiagen, Valenica CA, USA) following manufacturer's recommendations. Plasmid DNAs were purified from a minimum of three colonies for each reaction. To verify if cloning reaction was successful, PCR was performed using 1 μL of plasmid DNA as template and primers corresponding to those used to amplify the PCR fragment cloned into the vector. PCR fragments were analysed on agarose gel, ensuring the correct PCR fragments were obtained. Clones with correct PCR fragments were further confirmed by sequencing as described below. The concentration and purity of purified plasmid DNA was measured using NanoDrop 2000 (Thermo Fisher Scientific Inc, USA) following manufacturer's instructions. All DNA samples were required to have a 260/280 ratio of between 1.8 and 2.2. Most samples had 260/280 ratio of between 1.88 and 2.0. Plasmid DNA Genomic Equivalence (GE) was calculated using the following equation:

Bi-directional sequencing was performed either on plasmid DNA or directly on PCR fragments obtained from genomic DNA. DNA amplifications were performed as described above and PCR fragments purified using QIAquick PCR purification kit. BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Carlsbad CA, USA) was used for bi-directional sequencing following manufacturer's recommendations. Each reaction contained 4 μL of Terminator Ready Reaction Mix, 3.2 pmol primer, template DNA (plasmid DNA or purified PCR fragment) in a final volume of 10 μL. M13F or M13R primers which are part of TOPO TA cloning kit were used for the bi-directional sequencing of plasmid DNAs whereas forward or reverse primers used in generation of the PCR fragments were used in sequencing of these amplicons. Cycling conditions for the BigDye Terminator reaction was as follows: 96°C for 1 min followed by 25 cycles of 96°C for 10s, 50°C for 5s, and 60°C for 4 min. The products of these reactions were purified using Performa Spin Columns (EdgeBio, Gaithersburg MD, USA) following manufacturer's recommendation. Electrophoresis of these samples was performed using 3130XL Genetic Analyzer (Applied Biosystems, Carlsbad CA, USA) following manufacturer's recommendation.

To analyse the performance of the selected SNP assays, three P. falciparum strains, 3D7, 7G8 and K1 that differ in some of their genetic profiles at selected codons (Table 3) were used to generate plasmid DNAs containing PCR fragment of genes carrying codons of interest. Four PCR fragments that contained all the eleven SNPs were amplified from respective genes and cloned into TOPO TA vectors as described in the methods section. Primers used for amplification of these PCR fragments are shown on table 2. Amplicon sizes were as follows; the amplicon that carried pfdhfr codons A16V, S22N, N51I, C59R, S108N, and I164L was 526 bp; the amplicon that carried pfmdr1 codons N86Y and Y184F was 735 bp; the amplicon that carried pfmdr1 codons S1034C and N1042D was 447 bp; and the amplicon that carried pfdhps codons A581G was 357 bp. Pfdhfr codons A16V, S22N, N51I, C59R, S108N and I164L PCR fragments were amplified from 3D7 and 7G8 strains. These fragments were cloned into TOPO TA vectors and the plasmid DNAs generated are referred to as 16-3D7 and 16-7G8. Plasmid DNA 16-3D7 was used for analysis of codons A16V, S22N, C59R, and I164L. Note that the genetic profiles of these codons are the same (wild-type) in 3D7 and 7G8 strains. Plasmid DNAs 16-3D7 and 16-7G8 were used for analysis of codons N51I and S108N, which 3D7 and 7G8 strains have different genetic profiles at these codons. For codons pfmdr1 N86Y, Y184, S1034C, N1042D and pfdhps codon A581G, plasmid DNAs were generated by cloning PCR fragments from K1 and 7G8 strains. K1 and 7G8 strains have different genetic profiles at these codons. The generated plasmid DNAs are referred to as follows; plasmid DNAs carrying pfmdr1 codons N86Y or Y184F are referred as 86-K1 or 86-7G8; plasmid DNAs carrying pfmdr1 codons S1034C and N1042D are referred to as 1034-K1 or 1034-7G8; and plasmid DNAs carrying pfdhps codon A581G are referred to as 581-K1 or 581-7G8. The concentration of each plasmid DNA was determined using NanoDrop and the genomic equivalence (GE) was calculated. To obtain the initial analytical sensitivity and specificity, each plasmid DNA was diluted to an initial final concentration of 32, 000 copies. SNP assays were performed as described in methods section. SNP assay names were based on the codon being analysed. For example, to analyse codon pfdhfr N51I, the SNP Assay name is referenced as DHFR51 (Table 1). Performance of each SNP assay was assessed based on the known genetic profile of each strain at each codon and sequence data. Assessment of each SNP assay was based on automatic calls, normalized dye fluorescence (ΔRn) as a function of cycle number for Allele1 (wild-type) or Allele2 (mutant), and the CT values. All the calls made (11 of 11) for each SNP assay were correct, establishing the initial analytical sensitivity and specificity at 100%. To establish the Limit of Detection (LoD) for each assay, each plasmid DNA was serially diluted 5-fold and the initial tentative LoD for each assay was set as the lowest concentration of DNA that yielded positive results in duplicate experiments (Figure 1: shows LoD for pfdhfr 51 and 108, pfmdr1 86 and 184, pfmdr1 1034 and 1042, pfdhps 581). The LoD was confirmed by testing the LoD concentration in replicates of two or more. All the assays had LoD of 2 GE.

To test the ability of each assay to discriminate mixed infections, plasmid DNAs carrying alternate genetic profiles at each codon were mixed at different ratios and the ability of each assay to make the correct call was assessed. SNP assays analysed for mixed infection were DHFR51 and 108 (using plasmid DNA 16-3D7 and 16-7G8), MDR86 and 184 (using plasmid DNA 86-K1 and 86-7G8), MDR1034 and 1042 (using plasmid DNA 1034-K1 and 1034-7G8), and DHPS581 (using plasmid DNA 581-K1 and 581-7G8). Analysis for mixed infection was performed as follows; one plasmid DNA concentration was kept constant while the other was serially diluted 5-fold or both plasmid DNAs were serially diluted 5-fold and mixed at equal concentrations. As shown in Figure 2 (DHFR108 SNP assay, MDR1034 SNP assay and DHPS581 SNP assay), when one plasmid DNA concentration was kept constant (high) while the other was serially diluted, the presence of heterozygous alleles was only detected when both plasmid DNAs were at high concentration. The presence of the allele in the serially diluted plasmid DNA was not detected beyond when the plasmid DNAs were at equal concentrations. However, for MDR1034 SNP assay, both alleles were detected when one plasmid DNA was diluted to 6400 GE (Figure 2, MDR1034 SNP assay) while the other was at kept at 32000 GE. Curiously, the software did not make automatic calls for the DNA mixtures that contained serially diluted plasmid DNA at 1:5 (6400 GE) and undiluted plasmid DNA (32000 GE) for DHFR51, MDR1042 and DHPS581 SNP assays, rendering these six assays undetermined. There is no clear explanation why such results were obtained (undetermined) at these plasmid DNA mixture ratios since the software made automatic calls for the consequent plasmid DNA mixtures. Note that the software did not make automatic calls for all MDR86 and MDR184 SNP assays. Notwithstanding, Allele1 and Allele2 ΔRn in these SNP assays were clearly separated but yet the software could not make automatic calls (Figure 3). Manual calls were made based on ΔRn values of the Allele1 and Allele2. It was determined the performance of these two SNP assays, MDR86 and MDR184 were similar to those of the other SNP assays. Interestingly, in all the SNP assays, when both plasmid DNAs were serially diluted, the presence of both alleles was detected at or near LoD as established using single plasmid DNA.

Twelve samples from the clinical trial were randomly selected and sequenced directly or by cloning PCR fragment that contains codons of interest into TOPO TA vector. Primers used for cloning and/or sequencing are described in Table 2. Here, similar scheme was used as that described in the results section, SNP assay performance. PCR fragments were amplified using different primer sets and amplicons were either cloned into TOPO TA vector and then sequenced or were directly sequenced as described in the methods section. SNP assays were performed in replicates of two or more for each sample using either cloned plasmid DNA and/or genomic DNA of the clinical samples. 3D7, 7G8 or K1 plasmid DNAs were used as positive controls for each allele in all the assays analysed. For DHFR16, 22, 59, and 164 SNP assays, only 16-3D7 plasmid DNA was used as positive control (Allele1) because at these genetic loci, both 3D7 and 7G8 strains carry wild-type alleles. For DHFR51 and 108 SNP assays, 16-3D7 and 16-7G8 plasmid DNAs were used as positive controls (Allele1 and Allele2 respectively) because at these genetic loci, 3D7 and 7G8 carry different alleles (Table 3). 86-7G8 and 86-K1 plasmid DNAs were used as Allele1 and Allele2 positive controls respectively for MDR86 SNP assays whereas for MDR184 SNP assays, 86-K1 and 86-7G8 plasmid DNAs were used as Allele1 and Allele2 positive control respectively. 1034-K1 and 1034-7G8 plasmid DNAs were used as Allele1 and Allele2 positive controls for MDR1034 and 1042 SNP assays respectively. For DHPS581 SNP assay, 581-7G8 and 581-K1 plasmid DNAs were used as Allele1 and Allele2 positive controls respectively. K1 and 7G8 strains differ at all these genetic loci as shown (Table 3). SNP assays were performed as described in the methods section. Figure 4 shows an Allelic Discrimination plots for DHPS581 and MDR86 SNP assays ran using clinical samples and plasmid DNA controls. The plot shows a clear and distinct separation between the two alleles. In these experiments unlike those initially ran, automatic calls were made for MDR86 and 184 SNPs assays since large number of samples and controls were used, emphasizing the need of analyzing a large number of samples for the software to make automatic call. To assess the analytical specificity of the SNP assays, calls made using SNP assay data were compared to the sequence data of the samples. All calls made using SNP assays were in 100% agreement with sequence data. Sequence analysis is considered a gold standard method for SNP analysis. Eleven SNP assays were analysed in 12 clinical samples, representing a total of 132 SNP assays that were correctly called. This analysis, in addition to that initially performed using 3D7, 7G8 and K1 strains represent 100% specificity of all the SNP assays analysed.

Sixty clinical samples with known parasite density based on microscopy and real-time PCR data [27] were analysed using all the eleven SNP assays. All the SNP assays performed as expected with distinct and clear calls regardless of the parasite density in each sample. In additional, 10 clinical samples that were negative by microscopy and real-time PCR data were ran for further analysis of the analytical specificity of the SNP assays. There were no false positives; all the CT values were undetermined.