All experiments utilized the P. falciparum NF54 line cultivated at 5% haematocrit in RPMI 1640 medium, 0.5% Albumax II (Invitrogen, Carlsbad, California, United States), 0.25% sodium bicarbonate, and 0.1 mg/ml gentamicin. Parasites were incubated at 37°C in an atmosphere of 5% oxygen, 5% carbon dioxide, and 90% nitrogen. Parasites were transfected by using "DNA loaded" red blood cells as previously described12. Briefly, 0.2 cm electroporation cuvettes were loaded with 0.175 ml of erythrocytes and 50 μg of plasmid DNA in incomplete cytomix solution. Electroporation conditions were 0.31 kV and 960 mFD. For stable transfections, parasites were cultured in media containing either 40 ng/ml WR99210 or the designated concentration of blasticidin.

Construction of pHLIDH constructs: Plasmids pVLH and pVLHIDH were described previously13. The luciferase coding region and hrp2 3'UTR of pVLH were amplified by PCR introducing restriction sites for PstI (5') and KpnI (3') using the primers FflucPstI: 5'-AACTGCAGGCATGGAAGACGCCAAAAAC-3' and hrp3'downKpnI: 5'-GGGGTACCCGCCTCTCCCCGCGCGTTGG-3'. Subsequently, this fragment was ligated into PstI/KpnI digested pVLHIDH to replace the VLH portion (var promoter, luciferase and hrp2 3'UTR) to yield plasmid pHLIDH. In the resulting plasmid, the PFB1055c intron acts as a promoter for both the reporter and drug resistance genes. Additional intron sequences were first amplified from genomic DNA with the following primers, thereby adding the restriction sites BamHI and SalI to the 5' and 3'ends, respectively, and inserted into BamHI/SalI digested pHLIRH. Gene specific primers for var introns: PFC0005w: 5'-CGGGATCCGAAGGTAAAAGAGAATATATATGTG-3'/5'-GCGTCGACTTCTAAAATAATAAAAGAGG-3'; PFF0845c: 5'-CGGGATCCCTAAAGGTATTATATATG-3'/5'-GCGTCGACGATCAATAGTAGATTTGG-3'; PFD0020c: 5'-CGGGATCCGCGCTTTTATTTTTGAAGG-3'/5'-GCGTCGACGGTCCACAGGAGATTTAGG-3'. For the construction of pHLsti/CamDH, the intron in pHLIDH (SmaI and SalI) was replaced with the PfCam promoter from pHH-Tati2 construct (HpaI and SalI) (gift of M. Meissner)8.

Contruction of pHBIRH vectors: Luciferase gene was cut out pHLIRH with HindIII and SmaI and replaced with the blasticidin S deaminase gene obtained by restriction digestion of pVcBB/IDH14 with HindIII and HpaI. Restriction sites BamHI and SalI were used to exchange bidirectional promoters between plasmids pHLIDH, pHLIRH and pHBIRH.

Parasites were synchronized by magnetic cell separation using MACS CS columns (Miltenyi Biotech) and confirmed by light microscopy. Parasitaemias were counted for 1000 RBCs. For luciferase assays, parasites were obtained from 200 μl of culture by centrifugation and subsequent lysis in 100 μl of Glo Lysis Buffer (Promega). To determine firefly luciferase expression, 100 μl of Bright-Glo Luciferase reagent was added to the lysate and luminescence was measured immediately in a TD-20/20 luminometer. For Renilla luciferase expression, 100 μl Renilla assay reagent (Promega) was added to the lysate and luminescence determined using the TD-20/20 luminometer. For both Renilla and firefly luciferase, luminescence was expressed per 1% parasitaemia.

RNA was extracted from synchronized ring stage parasites 16–18 h or late trophozoites 40–42 h post-invasion. RNA extraction was performed with the TRIZOL LS Reagent (Invitrogen) as previously described15. RNA to be used for cDNA synthesis was purified on PureLink column (Invitrogen) according to manufacturer's protocol. Isolated RNA was then treated with Deoxyribonuclease I (Invitrogen) to degrade contaminating gDNA. cDNA synthesis was performed with Superscript II Rnase H reverse transcriptase (Invitrogen) as described by the manufacturer. cDNA was synthesized from 800 ng total RNA in a reaction volume of 20 μl. For each cDNA synthesis reaction, a control reaction without reverse transcriptase was performed with identical amounts of template and primers. Realtime PCR was carried out as described previously14. All reactions were performed at a final primer concentration of 0.5 μM using Bio-Rad ITAQ SYBR SUPERMIX^® in 20 μl reactions on an ABI Prism^® 7900HT real-time PCR machine.

All reactions included primers for the control genes seryl-tRNA synthetase (PF07_0073), fructose biphosphate aldolase (PF14_0425) and actin (PFL2215w), as described by Salanti et al16, as well as arginyl-tRNA synthetase (PFL0900c) and glutaminyl-tRNA synthetase (PF13_0170), as described by Dzikowski et al14. The ΔCT for each individual primer pair was determined by substracting the individual CT value from the CT value of the control the seryl-tRNA synthetase. ΔCTs were then converted to relative copy numbers with the formula 2^ΔCt. Primers for luciferase17 and blasticidin14 were previously described. Additional gene-specific primer pairs were designed using primer 3 software18: Renilla luciferase: 5'-TTCGAAAGTTTATGATCCAG-3'/5'-AACATGTCGCCATAAATAAG-3'

msp1: 5'-TACAAGTCCATCATCTCGTT-3'/5'-TGGTTAAATCAAAGAGTTCG-3'

kahrp: 5'-CAATTACAACCTCAACAACC-3'/5'-TTTTACCATCGACAACATTT-3'

sbp1: 5'-AATCCACAACTGATTTGGTA-3'/5'-GAATAGGGGACATAGATTCG-3'.

In an attempt to create smaller, more stable expression constructs for transfecting cultured P. falciparum parasites, plasmids were constructed that contained both the dhfr and luciferase coding regions arranged such that expression of both genes was driven simultaneously from a single bidirectional promoter (Figure 1A). Three different promoters were compared: two var introns from genes PFB1055c and PFC0005w, respectively, and the 5' upstream region of the Pfcam gene. The promoter activity of var introns has been previously documented13, and the ~1 kb region upstream of Pfcam had been previously shown to have bidirectional promoter activity3. In a more detailed analysis19 it was shown that this sequence is located between two genes in head-to-head orientation and contains two different promoters, one for the Pfcam gene and a second one in the reverse orientation for an open reading frame that shows homology to stress-inducible genes in yeast and hence has been named P. falciparum stress-inducible gene (Pfsti). P. falciparum parasites were transfected with all three constructs and, after selection with WR99210, stably transformed parasite lines were obtained with each of the constructs, indicating the dhfr gene was expressed at sufficient levels to readily produce drug resistant parasites.

To determine the efficiency of transgene expression from the different constructs, the three parasite lines were synchronized and luciferase and dhfr expression were analysed at different stages of the intraerythrocytic cell cycle (Figure 1). The construct with var intron PFC0005w produced the highest level of luciferase expression, although all three transformed lines were well above background. luciferase expression of all three parasite lines was significantly higher in late stages than in ring stages (Figure 1B). For the two intron promoters this was expected and corresponds to the previously reported activity of var intron promoters and synthesis of non-coding var RNAs20. The Psti/cam promoter is active in a more constitutive manner and shows a lower degree of stage specificity. Compared to var intron PFB1055c, the Psti/cam promoter becomes active earlier in the cell cycle with luciferase levels ~10 fold higher in ring stage. Its activity shows a slower increase over the course of the cell cycle to finally achieve maximum expression in schizonts with luciferase activity similar to that produced by the construct with the PFB1055c var intron.

To quantitatively analyse expression of the dhfr gene in the transfected lines, realtime RT-PCR was used to measure steady state mRNA levels. This analysis showed that dhfr and luciferase expression correlated in all constructs, consistently showing stronger expression in late stages (Figure 1C). This indicates that the promoter activities in the forward and reverse orientations are not regulated independently, but are strictly linked. This is particularly interesting for the promoter region of Psti/cam as it contains two promoters for different genes that might have been expected to be regulated separately. It has been hypothesized that the Psti upstream regulatory region does not contain a constitutive promoter, but rather might be inducible only in response to cellular stress19. An induction of Psti promoter activity was not observed under temperature stress (data not shown) and it is assumed that the promoter activity measured in this assay corresponds to the "normal" promoter activity of the Psti gene. However, the possibility that the Psti promoter is inducible under different conditions and that in this assay a low "stand-by" activity was measured cannot be excluded.

Transcription levels of luciferase and dhfr were compared to two single copy P. falciparum genes, skeleton binding protein 1 (sbp1) and merozoite surface protein 1 (msp1). The sbp1 gene is expressed early in the cell cycle, therefore, sbp1 RNA levels in ring stages are much higher than for luciferase, while in late stages, sbp1 expression is low compared to luciferase. Thus the expression of luciferase and sbp1 are inversely correlated. As a consequence the expression constructs described here would not be expected to be suitable for the expression of ring-specific genes. In contrast, msp1 is expressed later in the cell cycle. Although msp1 transcription is not exclusively restricted to trophozoites, but starts already in late ring stages, msp1 expression increases over the course of the cell cycle and RNA levels peak in late trophozoites2122. Due to the late-stage specific activity of the intron promoters, the expression pattern of luciferase corresponds more closely to that of msp-1. In schizont stage the expression levels of luciferase produced by the construct with the stronger intron promoter (var PFC0005w) are very similar to those of msp1. Therefore, the pHLIDH constructs might be suitable vectors for the expression of late stage genes such as msp1.

In stably transformed P. falciparum parasites, episomes are usually present as concatemers2324. This potentially enhances the expression of the resistance marker by providing transfected parasites with multiple copies of the resistance gene. It had previously been observed that episome copy number can vary when blasticidin-S-deaminase (bsd) was used as a selection marker, showing a direct correlation of drug concentration and the number of plasmid copies. In the expression constructs described here containing bidirectional var introns, expression of the resistance and reporter genes were strictly linked. It was therefore hypothesized that it might be possible to regulate expression levels of a reporter gene (or a transgene of interest) by varying blasticidin concentration in the culture media. For this purpose, constructs were made that contained bsd as a resistance marker and Renilla luciferase (pHBIRH) as reporter gene, both simultaneously expressed from a bidirectional promoter. Two different var introns, PFC0005w and PFD0020c, were used as promoters (Figure 2). These constructs were used to stably transform P. falciparum NF54 parasites using selection with 2 μg/ml blasticidin. Both parasite lines displayed the expected stage-specific expression pattern, with approximately five-fold higher Renilla luciferase expression in schizonts than in rings. pHBIRH with the PFC0005w intron produced much stronger Renilla luciferase expression than did the intron from PFD0020c. Therefore, pHBIRH containing var intron PFC0005w was used for further experiments.

P. falciparum parasites stably carrying pHBIRH (intron PFC0005w) were divided into identical cultures and grown in the presence of 2, 5, 10 and 20 μg/ml blasticidin for four weeks. In the cultures containing 10 and 20 μg/ml blasticidin, dead parasites were detected initially and growth of the cultures slowed down significantly. However, after five cycles (10 μg/ml) or eight to 10 cycles (20 μg/ml), cultures had recovered and parasite growth resumed at usual rates. All four cultures were synchronized and Renilla luciferase activity was measured at the ring stage (~16 h post invasion) and trophozoite stage (~40 h post invasion). As expected, Renilla luciferase expression was stage-specific with ~five-fold greater activity in trophozoites (Figure 3A). In both stages, a linear correlation of Renilla luciferase activity and blasticidin concentration was observed, demonstrating that regulation of an episomally expressed reporter gene is possible by varying the concentration of the drug used for selection.

To investigate if the increase in Renilla luciferase expression resulted from an increase in promoter activity or rather reflected an increase in the copy number of the expression construct, DNA was extracted from the different parasite lines and the copy numbers of the bsd and Renilla luciferase genes were determined by quantitative PCR and compared to several single-copy genes of P. falciparum. There was a 10 to 20-fold difference in copy numbers between cultures grown in the presence of 2 μg/ml and 20 μg/ml blasticidin. In all four cultures, the copy numbers of the expression constructs displayed a linear correlation with both blasticidin concentration and Renilla luciferase activity (Figure 3B), indicating that the increase in bsd and Renilla luciferase expression was not due to changes in promoter activity but rather to a change in copy number of the expression construct. Thus the plasmid copy number can be regulated over a 10-fold range by altering the drug concentration used for selection and in this way the expression of the transgene (in this case Renilla luciferase) can be co-regulated. Episomal plasmids are usually present as concatemeres of up to 15 copies2324 suggesting that changes in concatemer size are the most likely mechanism for copy number variation, however, it is also possible that an increase in copy number is caused by presence of multiple episomes per parasite.

A possible application of these constructs could be the expression of a gene of interest in place of Renilla luciferase to investigate its putative biological function. As the effect of over-expressing genes in Plasmodium depends on the timing and the strength of expression25, the expression levels of Renilla luciferase and bsd were compared to the expression levels of several endogenous P. falciparum genes. cDNA was prepared from trophozoites and subsequently analysed by quantitative PCR. As expression from intron promoters is restricted to late stages, we compared the expression levels to that of msp1, a gene that is similarly primarily expressed in late stages. The relative amount of bsd and Renilla luciferase RNA directly correlated with blasticidin concentration, corroborating the results of the Renilla luciferase activity measurement (Figure 3C). RNA levels of Renilla luciferase were lower than msp1 for blasticidin concentrations of 2 and 5 μg/ml, however parasites growing in the presence of 10 μg/ml showed nearly equal expression levels of Renilla luciferase and msp-1 (Renilla luciferase 92% of msp1). In parasites growing in the presence of 20 μg/ml, Renilla luciferase RNA levels were ~40% higher than msp1 RNA. Other genes that are ring-stage specific (kahrp and sbp1) were expressed at lower levels in late trophozoites and served as controls. These results show that with the novel expression constructs described here, gene expression can be regulated over an approximately 10-fold range and, as was exemplified with msp1, expression levels can be achieved that correspond to endogenous transcription levels. In addition, the effect of a co-expressed gene (or a truncated version thereof) can be analysed by choosing concentrations that result in mRNA amounts above or below the "normal" expression level.