TEXT

Top Abstract Text References

Stable genetic transformation has been achieved in three species of Plasmodium, the agent of malaria: Plasmodium falciparum, a human pathogen (16); Plasmodium berghei, which infects rodents (12); and Plasmodium knowlesi, which infects primates (11). One of the main applications of this new technology is the analysis of protein function via gene targeting. In Plasmodium, targeting constructs integrate exclusively into the haploid genome by homologous recombination, which greatly facilitates gene targeting procedures (6). The P. berghei rodent system offers the advantage that targeted clones (erythrocytic stages of the parasite) can be efficiently selected in a limited period of time, generating clones that retain the ability to undergo gametocytogenesis and the complete cycle in mosquitoes. It also permits in vivo analysis of liver infection by the sporozoite stage of the parasite. The model is, however, still limited by the paucity of genetic tools. So far, only one activity, resistance to pyrimethamine, can be used to select transformants.

The green fluorescent protein (GFP) from the jellyfish Aequorea victoria has become a marker of choice for studies on gene expression and protein localization in living cells (see reference 7 for a review). Among the GFP variants that have spectral characteristics suited to conventional fluorescein isothiocyanate (FITC) filter sets (i.e., with a shift in excitation maximal from 395 nm to 480 to 500 nm) (2, 3, 5), GFPmut2 displays an ~20-fold more intense fluorescence than that of wild-type (WT) GFP when excited at 488 nm and is more soluble (2). This variant has recently been used as a reporter in P. falciparum transformation, by expressing it from the untranslated regions (UTR) of the P. falciparum hrp3 and hrp2 genes (14). Erythrocyte (RBC) stages of the parasite transiently transformed with the construct were readily detectable by standard FITC microscopy.

Construction of the PyrFlu selectable marker and of PyrFlu-expressing P. berghei lines. We adapted GFPmut2 to the stable transformation of P. berghei. We constructed a cassette, named PyrFlu, which confers pyrimethamine resistance and directs a fluorescence signal via a single protein fusion. The pyrimethamine resistance cassette originally used for transforming RBC stages of P. berghei (12) consisted of a P. berghei mutant DHFR-TS gene expressed by 2.5 and 1 kb of its own 5' and 3' UTR, respectively. Here, we constructed by amplification cloning a derivative of the original cassette that contains a DHFR-TS-GFPmut2 fusion gene under the control of 2.5 kb of 5'- and 0.5 kb of 3'-UTR of P. berghei DHFR-TS (Fig. 1A). The encoded fusion protein consists of all residues of each partner except the starting Met of GFPmut2, contains a Leu-Lys-Ala tripeptide linking TS and GFPmut2, and ends with an Ala-Glu-Phe tripeptide.

View larger version (26K): [in this window] [in a new window]   FIG. 1.   (A) Construction of the PyrFlu marker. The marker was constructed by amplification and cloning of three DNA fragments. The proximal fragment corresponding to the 3' end of the P. berghei pyrimethamine resistance DHFR-TS gene was amplified from plasmid pMD204 (12) with primers Oneca1ter (sense; 5'-GGTGCTGAATATACAGATATGCATGAT-3') and Ohar1 (antisense; 5'-CGCTTAAGAGCTGCCATATCCATATTTATTTTATCGT-3' [with the AflII site underlined]). The distal fragment corresponding to ~0.5 kb of the 3' UTR of DHFR-TS was amplified from the same plasmid with primers Ohar4 (sense; 5'-GCGGAATTCTAATGTTCGTTTTTCTTATTTATATAT-3' [with the EcoRI site underlined]) and ONeca4 (antisense; 5'-GCGGGTACCGGATCCATCGAAATTGAAGGAAAAAACATCA-3' [with the BamHI site underlined]). The central fragment corresponding to the GFPmut2 coding sequence was amplified from plasmid pHRPGFPM2 (14) with primers Ohar2 (sense; 5'-GCGCTTAAGGCTAGTAAAGGAGAAGAACTTTTCACT-3' [with the AflII site underlined]) and Ohar3 (antisense; 5'-GCGGAATTCTGCTTTGTATAGTTCATCCATGCCATGT-3' [with the EcoRI site underlined]). The three fragments were then fused via the AflII and EcoRI sites, and the internal BglII-BamHI fragment of the reassembled locus was used to replace its counterpart in plasmid pMD204, yielding plasmid pPyrFlu. The nucleotide and amino acid (single letter code) sequences immediately flanking the GFPmut2 sequence are shown. Sequences in bold originate from DHFR-TS; boxed sequences constitute linker sequences, each containing a restriction site (italicized); and underlined sequences originate from GFPmut2. Wavy lines indicate the multicopy plasmid pBSKS (Stratagene, La Jolla, Calif.). Ap, ApaI; Af, AflII; B, BamHI; Bg, BglII; E1, EcoRI; H2, HincII; H3, HindIII; K, KpnI; N, NotI; S1, SacI; S2, SacII; Sp, SpeI; Xb, XbaI; Xh, XhoI. (B) Targeting the PyrFlu marker at the TRAP locus in P. berghei. Plasmid pPyrFlu-TRAP was obtained by cloning ~3 kb of TRAP targeting sequence as a BamHI-NotI fragment into plasmid pPyrFlu. The TRAP targeting sequence consists of the 3' part of TRAP (5') and ~1.5 kb of its 3' UTR (thick line). Prior to transformation into WT P. berghei, plasmid pPyrFlu-TRAP was linearized at the unique SpeI site located in the targeting sequence 250 bp from its 5' end. Homologous integration of plasmid pPyrFlu-TRAP at the TRAP locus creates the INT locus, in which the first TRAP copy is full-length and expressed. Also shown are plasmid pBSKS (wavy lines), TRAP promoter (arrows), and 3' sequences necessary for the normal expression of TRAP (circles).

Fluorescence of RBC stages of recombinant parasites. Clear signals were detected by fluorescence microscopy with a standard FITC filter setting in RBC infected with either EPI or INT parasites (Fig. 2A). Fluorescence displayed by EPI parasites was stronger than that displayed by INT parasites, which probably reflects a gene dosage effect due to the high number of copies of DHFR-TS-based episomes per P. berghei parasite (up to 20 per nucleus [13]) versus the single copy of the cassette integrated in INT parasites. With both populations, however, the proportion of fluorescent RBC was always found to match the parasitemia determined by Giemsa staining of blood smears. This indicated that all RBC stages of the parasites expressed a functional fusion protein. This confirms earlier reports of constitutive transcription of the P. berghei DHFR-TS gene during the asexual development of the parasite in RBC (13).

View larger version (70K): [in this window] [in a new window]   FIG. 2.   Fluorescence of INT and EPI parasites at different stages of the life cycle. (A) RBC stages of the parasites. Blood was collected from the tails of infected rats, and cells were washed once in phosphate-buffered saline and examined by phase microscopy (left panel) or fluorescence microscopy with FITC filter settings. Note the multinucleated intracellular forms of the parasite and extracellular merozoites. Magnification, ×32. (B) Oocysts present in mosquito midguts at day 7 p.f. Midguts were dissected out from infected mosquitoes in RPMI medium and examined by phase or fluorescence microscopy. Magnification, ×32. (C) EEF of the parasites developing in vivo in the rat liver or in vitro in HepG2 cultured cells. For in vivo studies, young Sprague-Dawley rats (~60 g) were injected with 25,000 salivary gland sporozoites collected at day 18 p.f. Rats were then sacrificed 42 h later, their livers were removed, and frozen sections were prepared as described previously (1). For in vitro studies, HepG2 cells were incubated with 25,000 salivary gland sporozoites collected at day 18 p.f. for 2 h, extracellular sporozoites were washed, and cells were incubated at 37°C for ~42 h (4). Cells were then fixed with methanol, stained with monoclonal antibody 2E6 directed against parasite heat shock protein 70 (10), and treated with goat anti-mouse immunoglobulin conjugated to horseradish peroxidase, or they were examined by FITC fluorescence. Magnification, ×32.

View larger version (23K): [in this window] [in a new window]   FIG. 3.   (A) FACS analysis of blood infected with various parasite clones. Infected blood was collected by tail bleeding, diluted in phosphate-buffered saline to the concentration of 106 cells/ml, and analyzed 2 h (WT, INT, and INCO) or 30 min (INT-30 min) after bleeding with a Becton Dickinson FACScan machine equipped with an Argon laser tuned at 488 nm. The log fluorescence intensity of 10,000 cells (sorting events) was plotted against the forward scatter. (B) Schematic representation of the WT and INT recombinant loci. The predicted sizes of the fragments liberated upon genomic digestion with the restriction enzyme BamHI are indicated (in kilobases). Also shown are the TRAP 5' and 3' UTR (thick lines) and plasmid pBSKS (thin lines). (C) Sorting of PyrFlu-expressing RBC stages of the parasite. The blood of a rat (parasitemia of 1%) infected with a 1:1 ratio of WT and INT parasites (MIX) was sorted with a Coulter Epics Elite sorter (Coulter, Miami, Fla.). Gates on fluorescent (F) and nonfluorescent (NF) cells were set in a plot of fluorescence versus the forward scatter. Fluorescent and nonfluorescent cells were collected separately from the sample until the sort count of fluorescent cells was ~3,000 (and ~650,000 nonfluorescent cells had been sorted). Cells were injected into two rats, and the parasite DNA was collected from each rat (at 1% parasitemia) and analyzed by Southern hybridization with a probe corresponding to the entire coding sequence of TRAP. The BamHI digestion differentiates WT from INT parasites that appear as one band of 9 kb or two bands of 16 and 4 kb, respectively.

Fluorescence of sporogonic stages of recombinant parasites. We then examined the sporogonic cycle of PyrFlu-expressing parasites in Anopheles stephensi mosquitoes fed on infected hamsters (which induce higher infection rates in mosquitoes than rats). Oocysts from INT or EPI clones examined at day 7 postfeeding (p.f.) were brightly fluorescent (Fig. 2B). In contrast, the level of fluorescence emitted by individual EPI or INT sporozoites was low, and isolated sporozoites released in mosquito tissues could not be detected based on fluorescence. This may reflect a decrease in the level of expression of the DHFR-TS gene at this stage of the parasite, since merozoites that also contain a single copy of the genome are nonetheless brightly fluorescent.

Concluding remarks. The PyrFlu cassette appears to be a useful addition to the P. berghei transformation system. The relatively small size of the bifunctional cassette (5 kb) is compatible with the generation of more complex targeting plasmids, exemplified by plasmid pPyrFlu-TRAP. One copy of the PyrFlu cassette is sufficient to allow the selection of RBC stages of P. berghei via resistance to pyrimethamine or by flow cytometry. Pyrimethamine selection alone frequently yields a proportion of spontaneous pyrimethamine-resistant mutants that appear like the WT by Southern hybridization. The concomitant fluorescent signal should then allow the further selection of targeted clones by flow cytometry. A particularly appealing feature of a GFP-based marker is its potential use in negative-selection procedures. It should be possible by FACS to select spontaneous intrachromosomal recombination events that occur in a clonal population of parasite integrants containing the PyrFlu cassette, which revert to a WT structure and a nonfluorescent phenotype. Such a two-step approach with PyrFlu plasmids could then be used to introduce subtle gene mutations in a final locus devoid of the exogenous sequence.