Development and application of a positive-negative selectable marker system for use in reverse genetics in Plasmodium

Abstract

A limitation of transfection of malaria parasites is the availability of only a low number of positive selectable markers for selection of transformed mutants. This is exacerbated for the rodent parasite Plasmodium berghei as selection of mutants is performed in vivo in laboratory rodents. We here report the development and application of a negative selection system based upon transgenic expression of a bifunctional protein (yFCU) combining yeast cytosine deaminase and uridyl phosphoribosyl transferase (UPRT) activity in P.berghei followed by in vivo selection with the prodrug 5-fluorocytosine (5-FC). The combination of yfcu and a positive selectable marker was used to first achieve positive selection of mutant parasites with a disrupted gene in a conventional manner. Thereafter through negative selection using 5-FC, mutants were selected where the disrupted gene had been restored to its original configuration as a result of the excision of the selectable markers from the genome through homologous recombination. This procedure was carried out for a Plasmodium gene (p48/45) encoding a protein involved in fertilization, the function of which had been previously implied through gene disruption alone. Such reversible recombination can therefore be employed for both the rapid analysis of the phenotype by targeted disruption of a gene and further associate phenotype and function by genotype restoration through the use of a single plasmid and a single positive selectable marker. Furthermore the negative selection system may also be adapted to facilitate other procedures such as 'Hit and Run' and 'vector recycling' which in principle will allow unlimited manipulation of a single parasite clone. This is the first demonstration of the general use of yFCU in combination with a positive selectable marker in reverse genetics approaches and it should be possible to adapt its use to many other biological systems.

Figure 1

Generation of P.berghei parasites expressing negative selectable marker genes. (A) The constructs for expression of the hsvtk, bcd and yfcu negative selectable marker genes under control of the ef-1αa promoter. These constructs contain the tgdhfr-ts as a positive selectable marker and sequences of the small subunit (ssu) gene of the d-rrna locus for targeted integration into the genome. hsvtk, herpes simplex virus thymidine kinase; bcd, bacterial cytosine deaminase; yfcu, fusion gene of the yeast genes cytosine deaminase and UPRT; amp, ampicillin resistance gene; red boxes, 3′pbdhfr-ts repeat; K, KspI; H, HindIII; E, EcoRI. (B) Part of the genomic locus of the d-rrna gene that is used for integration of the expression constructs. ets, external transcribed spacer; its, internal transcribed spacer; ssu, small subunit. Sizes of PCR fragments and restriction fragments are indicated between short and long arrows, respectively. PCR primers are specified by their L-number. (C) The d-rrna locus after integration of the expression constructs in parasites that are selected by positive selection with pyrimethamine. (D) The d-rrna locus of parasites after negative selection with GCV or 5-FC of the mutant parasites as shown in (C). Parasites can be selected in which (1) the complete construct has been excised as a recombination event between the homologous d-ssu-rrna sequences or (2) the negative selectable marker cassette has been excised from the integrated construct by a recombination event between the two 3′pbdhfr-ts sequences. (E) Correct integration of the expression constructs in the parasite lines expressing hsvTK, bCD and yFCU after positive selection as shown by PCR. tgdhfr-ts, amplification of tgdhfr-ts (primers L190/L191); left, 5′ integration in the d-rrna unit (primers L739/L635); right, 3′ integration in the d-rrna unit (primers L1662/L740); M, marker. (F) Southern analysis of EcoRI and HindIII digested genomic DNA of wt and transformed parasites showing integration of the constructs in the d-rrna locus. DNA of wt and hsvTK, bCD and yFCU expressing parasites was hybridized to the c/d-ets probe.

Figure 2

Drug sensitivity of mutant parasites expressing negative selectable markers and in vivo selection of parasites. (A) Left panel: In vitro sensitivity of wt parasites and hsvTK expressing parasites to GCV. Right panel: Sensitivity of wt parasites and parasites expressing bCD, yFCU, TgDHFR-TS-yFCU and hDHFR-yFCU to 5-FC. Sensitivity is defined as the fraction of ring forms that developed into mature schizonts during a 24 h culture period in the presence of different concentrations of the drugs relative to the no drugs control. In each experiment, parasite development was determined by counting 100–200 parasites in Giemsa stained slides and percentages shown are the mean of counting duplicate wells of each drug concentration. (B) Southern analysis of the d-rrna locus in parasites expressing hsvTK before and after in vivo selection in mice with 100 mg (i.p.; m1–m3) or 500 mg (s.c.; m4–m5, i.p.; m6–m7) GCV/kg/day. Selection resulted in parasite populations consisting of mixtures of parasites with a disrupted d-rrna locus (10.6 kb, see Figure 1C and D) and parasites with a restored wt locus (17.3 kb, see Figure 1D). EcoRI and HindIII digested genomic DNA of wt, hsvTK and m1–m7 parasites was hybridized to the c/d-ets probe. (C) Southern analysis of the d-rrna locus in parasites expressing yFCU before and after in vivo selection in mice with 5-FC (m1–m5), showing selection of parasites with a restored wt d-rrna locus (17.3 kb fragment, upper panel, see Figure 1D) or parasites (asterisked lanes) that still contain a disrupted locus (10.6 kb fragment, upper panel) but in which the yfcu gene is lost (4.6 kb, middle panel and no signal, lower panel). The reduced size (4.6 kb instead of 6.8 kb) of the fragment that hybridizes to the tgdhfr-ts probe indicates a recombination event between the two 3′pbdhfr-ts sequences resulting in the excision of the yfcu gene in these parasites (see Figure 1D). EcoRI and HindIII digested genomic DNA was hybridized to the c/d-ets probe (upper panel), tgdhfr-ts probe (middle panel) or yfcu probe (lower panel).

Figure 3

Generation and analysis of parasites with a disrupted and subsequently restored p48/45 locus. (A) Construct used to disrupt p48/45 containing both a positive (tgdhfr-ts) and a negative selectable marker (yfcu). R, EcoRV; amp, ampicillin resistance gene; red boxes, 3′pbdhfr-ts repeat. (B) The genomic p48/45 locus. Sizes of PCR fragments and restriction fragments are indicated between short and long arrows, respectively. PCR primers are specified by their L-number. X, XbaI. (C) The disrupted p48/45 locus (dis1) after integration of construct 1 in parasites that are selected by positive selection with pyrimethamine. (D) The p48/45 locus after negative selection with 5-FC of the mutant parasites as shown in (C). Parasites can be selected in which (1) the wt genotype is restored as a result of a recombinant event between the homologous p48/45 sequences or (2) the negative selectable marker cassette has been excised from the integrated construct by a recombination event between the two 3′pbdhfr-ts sequences. (E) Southern analysis of the p48/45 locus in dis1 parasites before and after (m1–m6) treatment of infected mice with 5-FC, showing selection of parasites with a restored p48/45 locus [7.4 kb band in m1, 2, 3, 5, 6)]. The 15.7 kb band in m4 indicates a recombination event between the two 3′pbdhfr-ts sequences (see Figure 3C and D) resulting in the excision of the yfcu gene in these parasites. Genomic DNA was digested with XbaI and hybridized to the 5′p48/45 probe. (F) PCR analysis of the p48/45 locus in dis1 parasites before and after negative selection, showing a disrupted p48/45 locus in parasite line dis1, a restored wt p48/45 genotype in mouse m5 and a disrupted p48/45 locus with the yfcu sequence removed in m4 (asterisked lanes). p48/45, amplification of p48/45 (primers L1909/L1821); left, 5′ integration in p48/45 (primers L1909/L635); right, 3′ integration in p48/45 (primers L1662/L1821); yfcu, amplification of yfcu (primers L1504/L1505); M, marker. (G) Transcription of p48/45 as analysed by northern blotting in wt parasites, in mutants with a disrupted p48/45 locus (dis1) and in parasites with a restored p48/45 locus (m5) after 5-FC selection. RNA of blood stages enriched in gametocytes was hybridized to a p48/45 probe (upper panel) or the gametocyte specific control probe p28 (lower panel). (H) The presence of P48/45 as analyzed by western blotting in wt parasites, in mutants with a disrupted p48/45 locus (dis1) and in parasites with a restored p48/45 locus (m5). Proteins extracted from blood stages enriched in gametocytes reacted with polyclonal antibodies raised against recombinant P48/45 (upper panel) or against recombinant protein of the gametocyte specific protein P47 (lower panel). (I) Gametes of mutant parasites with a disrupted p48/45 (dis1) fail to fertilize, resulting in the lack of ookinete formation and the presence of clusters of female unfertilized gametes in ookinete cultures. Parasites with a restored p48/45 locus (m5) produce normal ookinetes. Parasites were stained with Giemsa and photographed at 1000× magnification. (J) Ookinete production in vitro in wt parasites, mutants with a disrupted p48/45 locus (dis1) and parasites with a restored p48/45 gene (m5) after negative selection. Parasites with a restored p48/45 locus produce normal numbers of ookinetes as shown by the percentage of female gametes developing in mature ookinetes (ookinete conversion rate).

Figure 4

Generation of parasites with a disrupted and restored p48/45 locus using a fusion gene of the positive and negative selectable marker. (A) Construct used to disrupt p48/45 containing a fusion gene of the positive (hdhfr) and negative selectable marker (yfcu). R, EcoRV; amp, ampicillin resistance gene. (B) The disrupted p48/45 locus after integration of construct 2 in parasites that are selected by positive selection with WR99210. Sizes of PCR fragments and restriction fragments are indicated between short and long arrows, respectively. PCR primers are specified by their L-number. X, XbaI. (C) Correct integration of construct 2 in the p48/45 locus in parasite line dis2 as shown by PCR. Left, 5′ integration in p48/45 (primers L1909/L635); right, 3′ integration in p48/45 (primers L1662/L1821); p48/45, amplification of p48/45 (primers L1909/L1821); M, marker. (D) Southern analysis of the p48/45 locus in mutant parasites before (dis2) and after (m1–m9) treatment of infected mice with 5-FC showing that selection with 1 g 5-FC/kg/day (m1–m3) results in mixtures of parasites with a disrupted p48/45 locus and parasites with a restored wt locus. Increasing the selective pressure by 5-FC treatment twice a day (m4–m6) or by a second round of 5-FC-selection of parasite population m1 (m7–m9) leads to a further enrichment for parasites with a restored p48/45 genotype.