Selection of drug resistant mutants from random library of Plasmodium falciparum dihydrofolate reductase in Plasmodium berghei model

Background The prevalence of drug resistance amongst the human malaria Plasmodium species has most commonly been associated with genomic mutation within the parasites. This phenomenon necessitates evolutionary predictive studies of possible resistance mutations, which may occur when a new drug is introduced. Therefore, identification of possible new Plasmodium falciparum dihydrofolate reductase (PfDHFR) mutants that confer resistance to antifolate drugs is essential in the process of antifolate anti-malarial drug development. Methods A system to identify mutations in Pfdhfr gene that confer antifolate drug resistance using an animal Plasmodium parasite model was developed. By using error-prone PCR and Plasmodium transfection technologies, libraries of Pfdhfr mutant were generated and then episomally transfected to Plasmodium berghei parasites, from which pyrimethamine-resistant PfDHFR mutants were selected. Results The principal mutation found from this experiment was S108N, coincident with the first pyrimethamine-resistance mutation isolated from the field. A transgenic P. berghei, in which endogenous Pbdhfr allele was replaced with the mutant PfdhfrS108N, was generated and confirmed to have normal growth rate comparing to parental non-transgenic parasite and also confer resistance to pyrimethamine. Conclusion This study demonstrated the power of the transgenic P. berghei system to predict drug-resistant Pfdhfr mutations in an in vivo parasite/host setting. The system could be utilized for identification of possible novel drug-resistant mutants that could arise against new antifolate compounds and for prediction the evolution of resistance mutations.


Background
Plasmodium falciparum, the most virulent malaria pathogen species, is responsible for nearly 863,000 deaths in 2009 [1]. Malaria treatment is hampered by existence of only a limited number of drugs and the emergence of parasites resistant to most available antimalarial drugs, including chloroquine, pyrimethamine and proguanil. Therefore, there is an urgent need to search for affordable, effective and safe anti-malarials that can combat drug resistant parasites.
A bifunctional enzyme in the folate biosynthesis pathway, dihydrofolate reductase-thymidylate synthase (DHFR-TS) is a well-defined target of traditional antimalarial drugs such as pyrimethamine and cycloguanil [2]. DHFR catalyses the production of tetrahydrofolate from dihydrofolate while TS is in charge of transferring a methyl-group from N5, N10-methylene-tetrahydrofolate to dUMP thereby generating dTMP and tetrahydrofolate. Mutations in Pfdhfr gene, associated with the amino acid substitution at residues 51, 59, 108, and 164, have been found in the field with different levels of resistance to antifolate drugs [3,4]. These mutations are positioned around the enzyme active site. Thus, the accumulation of point mutations in Pfdhfr reduces the affinity of antifolate drugs such as pyrimethamine for the enzyme leading to drug resistance [5,6]. These correlations of drug resistance with mutations in Pfdhfr gene have been experimentally verified by comparison of mutant enzymes in transfected parasites [7]. Despite the emergence of resistant Pfdhfr mutants in malaria endemic areas, the elucidation and development of target based screening models [8,9] and the solution of the crystal structure of PfDHFR-TS [10] means that this enzyme is still an attractive target for drug development. Therefore, in order to develop new compounds against drug resistant parasites, prediction of possible future resistance mutations becomes a priority.
Studies of mutant PfDHFR-TS expression in non-Plasmodium surrogate systems, such as Escherichia coli and Saccharomyces cerevisiae have successfully identified novel mutant Pfdhfr alleles that confer antifolate drug resistance [8,11]. These systems have also been useful in testing new compounds against the enzyme [8,9]. However, the biochemical activities of some mutant enzymes isolated using these systems have been shown to be significantly lower than the wild-type enzyme [11]. The physiology of these surrogate cells are markedly different from Plasmodium parasites, and it is not clear whether these resistant enzymes would have sufficient enzymatic activity to support parasite growth throughout all stages of the life cycle. Unfortunately owing to the narrow primate host range, suitable Plasmodium in vivo models are not available.
Rodent malaria parasites such as Plasmodium berghei are attractive models for human malaria, and the recent advances in DNA transfection technology allow for genetic modification of the parasite [12,13]. DNA transfection can be used to express P. falciparum genes in P. berghei, and thus transgenic P. berghei parasites can be used as in vivo surrogate models. In this study, the power of the transgenic P. berghei system to predict drug-resistant Pfdhfr mutations in an in vivo parasite/ host setting was demonstrated. The objective was to identify mutations in Pfdhfr that confer resistance to anti-folate drugs using a transgenic P. berghei model. Libraries of randomly mutated Pfdhfr were generated by PCR mutagenesis and transfected to P. berghei. Upon transfection, resistant parasites were obtained after pyrimethamine selection and found to harbor Pfdhfr mutant alleles. Thus, this system can be used as a Plasmodium surrogate system for more accurate prediction and identification of antifolate-resistance mutations.

Experimental animals and parasite
For all experiments, female BALB/c mice (National Laboratory Animal Center, Mahidol University, Thailand) 4-6 weeks old and weighing 20-25 g were used for P. berghei parasite infections. The transgenic P. berghei parasite line MRA-867 expressing green fluorescent protein without drug-resistant selectable marker (PbGFP), kindly provided by Drs. Andrew Waters and Chris Janse of Malaria Research Group, Leiden University Medical Center, the Netherlands, was used in this study [14]. All animal work was evaluated and approved by the Ethical Committee on Animal Experimentation, National Center for Genetic Engineering and Biotechnology (BIO-TEC), Thailand, and followed international guidelines for the use of animals in experimental studies.

Construction of P. berghei transfection plasmid
Plasmid for P. berghei transfection in this study was modified from the original plasmid pL0017 [15], which was kindly provided by Drs. Andrew Waters and Chris Janse (Leiden University Medical Centre, the Netherlands). The final transfection plasmid, designated pY005, contains wild type Pfdhfr-ts gene flanked with 1.0-kb each of 5' and 3' untranslated region (UTR) sequences of Pbdhfr-ts. BamHI and AflII restriction sites were introduced at 5' and 3' ends, respectively, of Pfdhfr domain to serve as cloning sites for the randomly-mutated Pfdhfr library.

Construction of Pfdhfr random mutant library
Pfdhfr mutant library was generated by error prone PCR [11]. The 50 μL PCR reaction contained 1 ng of pY005 plasmid template harboring wild type Pfdhfr, 10 μM of sense primer F1 (CGGTGGATCCATGATGGAACAAG; BamHI site is underlined), 10 μM of antisense primer R1 (CTTTGTCATCATTCTTAAGAGGC; AflII site is underlined), 0.1 mM dGTP, 0.1 mM dATP, 0.5 mM dCTP, 0.5 mM dTTP, 1× Mutagenesis buffer [16] and 5 units of GoTaq ® DNA polymerase (Promega). The thermocycle condition was: 1 cycle of 95°C for 3 min, 30 cycles of 95°C for 1 min, 50°C for 1 min, 72°C for 1 min, and final extension of 72°C for 5 min. PCR products of random mutant Pfdhfr library of about 0.7 kb were cloned into the BamHI/AflII sites of pY005. The plasmids containing Pfdhfr mutation libraries were grown in Luria Bertani broth containing 100 μg/ml ampicillin in a 37°C incubator shaker for 12-16 hours. Plasmids were extracted and purified using a Qiaprep Spin Miniprep kit (Qiagen). Extracted plasmids were precipitated by isopropanol and re-suspended in 10 μl TE buffer (10 mM Tris, 1 mM EDTA pH 8.0) for use in transfection experiment.

Determination of pyrimethamine sensitivity to parental and transgenic parasites
The sensitivity of pyrimethamine to inhibit the parental PbGFP parasite or transgenic parasites expressing PfDHFR mutant was determined by the 4-day suppressive test [17]. Five groups of five BALB/c mice per group were infected intravenously (i.v.) with 1 × 10 7 parasitized erythrocytes and treated with different concentration of pyrimethamine by intraperitoneal (i.p.) injection four hours post infection. The control group was treated with 5% (v/v) DMSO in PBS pH 4.0. The experimental groups were treated with different doses of pyrimethamine through the same route for another 3 days. Twenty-four hours after the last treatment (day 4), percentages of parasitaemia were determined by microscopic counting of Giemsa-stained smears from mouse tail blood.

Statistical analysis
All statistical analyses were carried out using SigmaPlot software version 11 (Systat Software Inc., USA). For calculation of the growth inhibitory curve, parasitaemia of the control group was set as 0% inhibition. The non-linear regression for sigmoidal dose-response (variable slope) was used to calculate the 50%, 90% or 95% effective dose (ED 50 , ED 90 or ED 95 ) values.

Transfection, selection and identification of Pfdhfr random mutant libraries
In vitro culture of PbGFP and P. berghei transfection were performed as described [13]. Briefly, parasitized blood was collected from a donor animal and cultured overnight in culture media (RPMI 1640 medium containing 20% heat inactivated fetal calf serum, 50 IU/mL neomycin and 25 mM Hepes). Schizont stage parasites were purified from the overnight culture by Nycodenz gradient centrifugation. The merozoites were transfected with the circular plasmid DNA harboring Pfdhfr mutant libraries using the standard Amaxa Nucleofector protocol [13] and re-infected into animals by i.v. injection. Twentyfour hours after transfection, 0.25 mg/kg of pyrimethamine was used to treat the infected mice by i.p. injection daily. Smears were taken to check parasitaemia and when positive, left to multiply, until parasite numbers were adequate (about 3% parasitaemia) for genomic analysis. Tail blood was drawn from infected animals on alternate days until parasitaemia reached 8-10% and genomic DNA was extracted from whole blood using a genomic DNA Mini Kit (Geneaid). The genomic DNA obtained was transformed into E. coli DH5α to recover the circular plasmid DNA containing Pfdhfr mutants. The transformed bacterial colonies were picked and extracted for plasmid DNA using a Qiaprep Spin Miniprep kit (Qiagen). The sequences of Pfdhfr mutants were obtained by DNA sequencing (Biodesign Sequencing Service, Thailand).
Transfection, selection and cloning of transgenic P. berghei parasite Plasmid containing resistant mutant Pfdhfr S108N was completely digested with HindIII and KasI restriction enzymes to generate linear plasmid. The 5' and 3'UTRs of Pbdhfr-ts served as homologous recombination sites for replacement of the endogenous Pbdhfr-ts on chromosome 7 with Pfdhfr S108N . In vitro culture of PbGFP and P. berghei transfection were performed as described [13]. Twenty-four hours after transfection, 0.25 mg/kg of pyrimethamine was used to treat the infected mice by i.p. injection daily until the drug resistant parasites appeared. The integrated transgenic mutant parasite clones were obtained by the limiting dilution method [18].

PCR analysis of transgenic P. berghei parasite
The correct integration of Pfdhfr S108N sequence into the genome via the 5' and 3'UTRs of the Pbdhfr-ts locus was determined by PCR. A 4.0-kb DNA fragment spanning the endogenous 5'UTR Pbdhfr-ts sequence and the introduced Pfdhfr was indicative of an integration event and was amplified using primers F2 (TTGAGCTACA-TAACTTCCATACAT) and R1 (described above). A 3.0kb DNA fragment spanning the introduced Pfdhfr and the endogenous 3'UTR Pbdhfr-ts sequence, indicative of a 3' integration event was amplified using primers F1 (described above) and R2 (CGATCTACACCTCTTCAT).

Expression profile of transgenic mutant parasite
The transgenic parasites expressed PfDHFR S108N -TS under the control of 5' and 3'UTRs of Pbdhfr-ts. To determine whether this promoter efficiently drives Pfdhfr S108N -ts mRNA expression, reverse-transcription PCR (RT-PCR) was performed. Total RNA was isolated from blood stage transgenic P. berghei parasites using Trizol reagent (Invitrogen). cDNA was generated and used as template for amplification with gene specific primers for the Pfdhfr transgene using primers F1 and R1 (described above). Controls included specific primers for Pbdhfr gene: PbDTF (GGGGGGGGCATATGGAAGACTTATCT-GAAACATTCG) and PbDTR (GGACTAGTGTACTT CCTCATTTGG) and P. berghei alpha tubulin gene: PbatubulinF (GCATGCTGGGAGCTATTTTG) and Pba-tubulinR (GCTGGTTCAAATGCTGAGTTTG). RT-PCR was performed using the same PCR condition as described above.

Determination of pyrimethamine sensitivity in wild type PbGFP parasites
Experimental mice infected with wild type PbGFP parasites were treated with different concentrations of pyrimethamine. As shown in Figure 1, in vivo ED 50 and ED 95 of pyrimethamine against PbGFP was 0.02 mg/kg and 0.25 mg/kg, respectively. The ED 95 concentration of pyrimethamine was then used for selection of transgenic resistant parasites in subsequent experiments.

Construction of Pfdhfr random mutant library using wild type Pfdhfr template
Approximately 14,000 bacterial colonies were obtained for the Pfdhfr random mutant library. Twelve colonies were randomly picked and sequenced. From the sequence alignment, up to 4 base substitutions per gene were found. The mutation frequency was then calculated as 0.26%, which is equivalent to approximately 2 base substitutions per 700 bp of Pfdhfr gene. Thus, the PCRinduced mutagenesis for this library is within the expected mutation frequency of 2 to 5 base substitutions per gene [19]. The rest of the bacterial colonies were harvested for library DNA preparation.

Selection of transfected resistant mutant parasites from wild type Pfdhfr random libraries
Plasmid DNA was purified from the Pfdhfr random mutant library and transfected to PbGFP parasites. Pyrimethamine-resistant parasites began to appear in most transfection experiments within 11 days post-transfection. Genomic DNA containing episomal transgenic DNA was extracted from drug-resistant parasites and transformed into E. coli. Among the isolated plasmids, two independent sequences were identified. One sequence contained a single base substitution mutation at amino acid position 108 (serine, S; AGC changed to asparagine, N; AAC) while the other had the same mutation at amino acid position 108 together with another base substitution at codon 196, which was silent (phenylalanine; TTT to TTC).
Generation of transgenic P. berghei expressing Pfdhfr S108N -ts mutant In order to confirm that the S108N mutation found among library-transfected DNA from drug-resistant parasites conferred resistance, allelic replacement of the endogenous Pbdhfr-ts with mutant Pfdhfr S108N -ts was performed. The strategy for the allelic replacement event is shown in Figure 2. Correct integration was investigated by PCR analysis on genomic DNA using different primers pairs. PCR products obtained corresponded with the expected 4.0 kb and 3.0 kb bands for 5' and 3'UTR integration respectively, as shown in Figure 3A.
Expression profile analysis of transgenic Pfdhfr S108N mutant parasite Transgenic P. berghei parasite with integrated Pfdhfr S108N -ts, named PbPfS108N, expresses the gene under the control of Pbdhfr-ts 5' and 3'UTRs. The expression of the enzyme was verified at the mRNA transcription level using RT-PCR. As shown in Figure 3B, 0.7 kb of Pfdhfr was amplified in cDNA from transgenic PbPfS108N parasite, but not in cDNA from PbGFP parental parasite. Conversely, Pbdhfr was amplified only in cDNA from parental PbGFP parasite, but not in cDNA from transgenic PbPfS108N parasites. The expected 0.85 kb band for the control P. berghei alpha tubulin was obtained for all parasites. No products were detected for -RT control templates, indicating that genomic DNA was absent.
Determination of growth rate and pyrimethamine sensitivity in transgenic PbPfS108N parasite The growth rate of transgenic PbPfS108N parasite was compared with PbGFP parental parasite. The growth rates of the two parasites were not significantly different ( Figure 4A). The efficacy of pyrimethamine to inhibit transgenic PbPfS108N parasites was determined by 4-day suppressive test. The calculated ED 50 and ED 90 values of  pyrimethamine against transgenic PbPfS108N parasites were 1.33 mg/kg and 2.65 mg/kg, respectively ( Figure 4B).

Discussion
There is need to identify possible drug-resistance mutations against new compounds under development as antimalarial drugs. Prior knowledge of possible resistance pathways facilitates testing of strategies for forestalling the evolution and spread of resistance in the parasite population, e.g. drug combinations. In this study, a P. berghei in vivo model was developed as a surrogate cell expression system to identify Pfdhfr mutations that confer antifolate drug resistance. This system employs DNA transfection in which the selectable marker is antifolate-resistant mutant Pfdhfr, rather than the conventional human dhfr and Toxoplasma gondii dhfr markers used in P. berghei transfection [20]. The concentration of pyrimethamine at 0.25 mg/kg (ED 95 of wild type P. berghei parasite) i.p. injection daily was successful in selecting resistant parasites obtained after transfection with Pfdhfr mutant libraries. Interestingly, the resistant parasite lines transfected with a Pfdhfr variant library were found to contain the S108N mutation, which is known to be the key antifolate resistance mutation found in nature [21]. The fact that the pyrimethamine resistant line carried this mutation provides proof of concept of this system to be able to identify drug resistant mutations from a mutant library.
PfDHFR-TS crystal structure reveals the asparagine side chain of the mutant enzyme is in steric clash with the pchlorophenyl moiety of pyrimethamine, which reduces the binding affinity of pyrimethamine for the mutant enzyme [10]. Although the selected drug-resistant library mutants contained the expressed S108N mutation, it should be noted that the drug-resistant library mutants are maintained episomally, and the increased dhfr gene copy number in episomes may contribute to drug resistance in  transgenic P. berghei, as has been shown for the hdhfr transfection marker [20]. Thus, to test the hypothesis that pyrimethamine resistance in transgenic P. berghei is determined solely by the S108N mutant allele, Pbdhfr-ts was replaced with the Pfdhfr S108N -ts mutant by double homologous recombination. The integrated transgenic parasites expressed PfDHFR S108N -TS with a single copy of mutant Pfdhfr S108N under the control of endogenous Pbdhfr-ts 5' and 3'UTRs. The resulting PbPfS108N transgenic parasites showed susceptibility to pyrimethamine with ED 50 values 66-fold higher than PbGFP wild type, which strongly suggest that resistance is conferred by the S108N mutation. In addition, PbPfS108N transgenic parasites grew the same as PbGFP parasites. This result demonstrated that the DHFR-TS function is conserved between the two Plasmodium species, in agreement with other cross-species comparisons in the same genus [22]. Furthermore, the pyrimethamine resistance mutation had no negative effect on the function of the enzyme, in agreement with earlier studies [21,23]. This demonstrates the capacity of the transgenic P. berghei parasite model to study the fitness of Pfdhfr mutant alleles in the whole transmission cycle.
For pyrimethamine resistance, in addition to S108N mutation, other mutations such as N51I, C59R and I164L are known to augment the resistance level in parasites [21,23]. Although our mutant library is of sufficient diversity to contain these other mutations, they were not found among drug-selected parasites. In order to obtain transgenic parasites with a higher level of resistance, higher drug pressure will needed. However, selection of highly resistant pyrimethamine alleles with two or more mutations is not feasible, since mice could not tolerate the very high pyrimethamine doses needed.
In conclusion, a new experimental system for predicting the evolutionary pathway of antifolate drug resistance was developed. The major advantage of this system is that drug-resistant mutant alleles can be selected from diverse Pfdhfr libraries in a Plasmodium surrogate cell in a proper host background. Proof of concept for the system was demonstrated using the well-known antifolate pyrimethamine; however, new antifolates are being developed which are also effective against pyrimethamine-resistant parasites [24]. Therefore, the system could be utilized for identification of possible novel drug-resistant mutants that could arise against new antifolate compounds. This information could be used for rational design of effective anti-malarial drugs that forestall the emergence of drug resistance. Furthermore, our approach could also be applied to other Plasmodium enzyme drug targets for prediction the evolution of resistance mutations.