Parasites were isolated from Dutch volunteers exposed to mosquitoes infected with P. falciparum isolate NF5418 as part of ongoing studies of experimental P. falciparum infections. On day 0, ten non-immune volunteers were subjected to two or five infectious bites. Chloroquine treatment was initiated on the first day a thick smear was positive. Parasite cultures were established from 400 microlitres of packed blood cells drawn on days 8, 9, and 10 and parasites were cultured in vitro for 27 or 33 days (Figure 1) to obtain sufficient parasite material for DNA, RNA and protein analysis. The parasites were cultured in 0 Rh⁺ erythrocytes as described 19, with the addition of 2% non-immune human serum to the culture media. Long-term in vitro 3D7 cultures expressing VSA_UM-type antigens or selected in vitro to express VSA_SM-type antigens were used as controls20.

The development of parasitaemia was monitored by quantitative real-time PCR as described 21. DNA, RNA and cDNA for var gene transcription analysis were prepared from synchronized parasite cultures as described 17. Quantitative real-time PCR was performed using a Rotorgene thermal cycler system (Corbett Research, Motlake, Australia). Real-time PCR-optimized and gene-specific primers for each of all full-length var genes and a pseudogene in the 3D7 isogenic NF54 P. falciparum genome were those described in 22, except for PFI1830c. Real time primers for this gene were forward 5'ACAACAATTTCGCAAGCAAG 3', reverse 5'TTCCTCTGCCTCCTCTTCAT 3'. Standard curves for the estimation of product-related fluorescent bias and amplification efficiencies were generated for all primer pairs. For 15 primer sets, standard curves were generated both from dilution series of genomic DNA and from cloned gene fragments 17. As the two approaches led to identical standard curves (not shown), standard curves for the remaining genes were determined from genomic DNA only. The standard curves were linear across a range of seven logs of DNA concentrations (R = 0.9779 to 0.9995) with amplification efficiencies between 90 and 101%. Standard curves were used for primer bias corrections in calculations of absolute transcript levels. The detection limit of the system was ≥ 20 copies. The housekeeping genes seryl-tRNA synthetase and fructose-bisphosphate aldolase have uniform transcription profiles throughout the parasite life cycle22, and were used as endogenous controls. Differences in var transcript distribution between samples were calculated by the ΔΔ CT method using the endogenous controls for normalization.

Immunoblot analysis was performed as previously described 17. Flow cytometryand plasma from malaria-exposed children and adults were used to classify the VSA expressed by parasites isolated on days 8, 9 and 10 as previously described72023

Successful infections were established in eight of the 10 volunteers. Despite low parasitaemias, a total of 13 parasite culture lines from blood collected from six of the volunteers at three time points after infection were established (Figure 1A). The highest parasite growth rates were observed in volunteers 1 and 2 with peak parasitaemias of 13,832 parasites/ml and 4,637 parasites/ml, respectively, whereas peak parasitaemias in all other volunteers were below 1,400 parasites/ml. As recognized from previous similar human vaccination trials 21, parasitaemias fluctuated in a distinct manner, probably reflecting liver release, sequestration of trophozoite/schizonts and release of new generations of merozoites from schizonts. Parasites obtained on day 8 were predominantly first-generation blood parasites, assuming that parasites were released from the liver between days 6,33 and 7,3324 (Figure 1B). Similarly, day9 parasites were assumed to represent second-generation parasites, and day10 parasites a mixture of second and third generation parasites.

The pattern of var gene transcription was surprisingly consistent in all the six parasite lines obtained on day8 from six of the volunteers (Figure 2). Transcripts of all var genes could be detected in ring-stage cultures, and most were transcribed at roughly similar levels. Interestingly, nine of the 10 lowest transcribed genes belonged to var subgroup A or B/A, which have previously been associated with severe malaria 13. In agreement with previous studies 17 all var genes were transcribed at markedly lower levels in thetrophozoite/schizont-stage compared to ring-stage parasites (data not shown). The pseudo-gene PFE1640w (var1) behaved differently and was expressed at similar levels by late stage parasites, comprising approximately 20% of the total number of var gene transcripts in trophozoite-stage parasites (data not shown).

The var gene transcription patterns of the five isolates obtained on day9 and the two obtained at day10 were different from those of the day8 isolates, in particular in the isolates obtained from volunteers 1 and 2 in whom high parasite growth was observed (Figure 1 and 3). Ring-stage parasites from volunteer 1 showed remarkably large transcriptional changes for five var genes. Four (PF11_0008, PFD1235w, MAL6P1.1, MAL7P1.55) were transcribed at much higher (>20-fold) and one gene (PFA0015c) at markedly lower levels (12-fold) in the day10 isolate compared to the day8 isolate. The estimation of absolute copy numbers (not shown) revealed that PF11_0008, MAL6P1.1 and MAL7P1.55 were the three highest transcribed var genes in the day10 isolate, comprising approximately 22%, 40% and 8% of all var transcripts, respectively. The abovementioned five genes also showed the most pronounced changes in gene transcription (10–20 fold) when trophozoite stage parasites from days 8 and 10 were compared. The most prominent change among ring-stage parasites from volunteer 2 was the 15-fold increased transcription of PFD0020c, which was the highest (~8%) transcribed var gene in the day 9 isolate (Figure 3). Only minor changes in var gene transcription patterns between days 8 and 9/10 were observed in the parasites isolated from volunteer 4, 6 and 10.

Overall, none of the var genes had a consistently altered transcript proportion in all volunteers. However, transcription of seven genes (PFD0005w, PFD0020c, PFD1235w, MAL6P1.1, PF10_0406, PF11_0007 and PFL0005w) increased in more than one volunteer, whereas transcription of three genes (PFA0015c, MAL6P1.4 and PFE1640w) decreased in more than one volunteer (Figure 3). In general, analysis of ring-stage and trophozoite/schizont-stage parasites yielded similar results.

To investigate PfEMP1 translation in the 13 isolates, western blot analysis were performed on protein extractions of trophozoite/schizont-stage cultures using rabbit and murine antisera raised against the conserved acidic terminal segment (ATS) and against DBL5δ of PFD1235w, respectively (Figure 4). The analysis of parasites from volunteer 1 revealed differential expression from day8 to day10 (Figure 4A). Thus, a high molecular weight band of around 400 kDa was seen only in the day 10 culture when using the ATS-specific antibody (Figure 4A). This band size corresponds to the expected size of PFD1235w, which also exhibited a large increase in transcription from day8 to day10 among parasites from this donor, and its identity was confirmed using the PFD1235w DBL5δ antibody (Figure 4B). PFD1235w could not be detected in any of the other cultures, although small increases in transcription were found in several of the cultures (not shown). Two bands of 350 kDa and 260 kDa were particularly intense in the day10 culture of volunteer 1 (Figure 4A). This finding correlates with the expected sizes and elevated transcription of PF11_0008 (345 kDa), MAL6P1.1 and MAL7P1.55 (both ~258 kDa). Transient bands at around 230 kDa and 310 kDa appeared to emerge on day 9 and disappear on day 10 in the cultures of volunteer 1. No obvious identity could be assigned to these proteins. In the remaining cultures, differential PfEMP1 expression detected by the ATS antibody was found in volunteers 6 and 10. In both cases a band of 310 kDa was observed in day10 and 9 parasites, respectively. The best candidate gene for this band is PFD0005c, which is predicted to have the observed size and was found to be more highly transcribed in the these cultures. All day 8 cultures were very similar and all appeared to dominantly express PfEMP1s around 260 kDa, which corresponds to the expected sizes of the highest transcribed genes in these cultures (figure 2).

The VSA phenotype of parasites can be classified relative to each other and in the spectrum between VSA_SM and VSA_UM, depending on the proportion of individuals from an endemic area that possess antibodies to the expressed VSA 720. To establish their VSA phenotype, the serological recognition of the isolated NF54 parasites was tested using a panel of plasma obtained from children and adults living in Coastal Ghana (Figure 5). All day 8 parasite isolates, representing the first generation of asexual blood-stage parasites, expressed VSA that were recognized by IgG in plasma samples from only a minority of the children and from about half of the adults (Figure 5). These parasites were also less well recognized than the standard 3D7 VSA_UM line, which dominantly expresses PfEMP1 encoded by GroupC var genes. The seven lines isolated on days 9 or 10 all expressed VSA that were more frequently recognized than the corresponding day8 line (P = 0.01, Wilcoxon signed-ranked test). This trend was particularly clear for the day9/10 lines from volunteers 1, 2 and 10, which were recognized by all of the plasma samples from adults and from the majority of the children, similarly to the recognition of 3D7 selected for expression of VSA_SM-type IE surface antigens (Figure 5). However, quantitatively the serum recognition of all isolates was lower than that of the 3D7_SM line (data not shown). The 3D7 VSA_SM line has been selected for VSA_SM expression using IgG from semi-immune children and dominantly expresses the product of PFD1235w (VAR4) on the surface on infected erythrocytes. Interestingly, parasites from volunteers 1, 2 and 10 had an increased transcription of a particular var gene (PFD1235w; var4) (Figure 2) as does 3D7 in response to selection for VSA_SM expression17.