Background
The most virulent form of malaria is caused by
Plasmodium falciparum [
1]. The virulence of
P. falciparum is a consequence of adhesion of infected red blood cells (iRBCs) to different host endothelial receptors [
2]. This process is primarily mediated by a polymorphic protein family collectively referred to as
P. falciparum membrane protein 1 (PfEMP1) [
2‐
4]. Expression of PfEMP1 variants that exhibit binding to chondroitin sulfate [
5,
6] in the placenta or to the endothelial protein C receptor (EPCR) [
7] has been associated with the development of placental malaria and cerebral malaria, respectively. PfEMP1 is encoded by the
var gene family [
8], a multicopy gene family with approximately 59–60 different genes per parasite genome [
9]. Only one
var gene is expressed in an individual parasite at a time, a process referred to as mutually exclusive expression [
10]. Switching between the active
var locus generates antigenic variation leading to immune escape of the parasites [
11].
var genes can be grouped according to their distribution across the 14
P. falciparum chromosomes into subtelomeric or central and this position correlates with their promoter type [
12]. 36 of the 59
var genes in the 3D7 genome strain are located in subtelomeric areas and have upstream (Ups)A, UpsB or UpsB/A type promoters, while 23 are located in central
var clusters and have UpsC or UpsB/C type promoters. An individual
var gene typically consists of exon 1 and exon 2 that encode for the extracellular and intracellular parts of the PfEMP1 protein, respectively. Exon 1 is hypervariable and consists of an N-terminal segment, followed by a variable number of domain cassettes (DCs) that are composed of several Duffy Binding Like- (DBL) and Cysteine Rich Interdomain Region-(CIDR) domains [
9]. DBL and CIDR domains are assembled of conserved sequence blocks that share similarities in all parasite strains as well as hypervariable sequence blocks [
13]. The 60
var genes of an individual parasite are almost completely distinct from each other and different
P. falciparum strains carry almost completely different
var gene repertoires [
14,
15]. This tremendous genetic diversity appears to be the result of the highly recombinogenic nature of the AT-rich
P. falciparum genome, the diversifying pressure of the immune system and the virtual absence of purifying selection on the
var gene family. Indeed recombination events generate new
var gene variants during meiosis and mitosis [
16,
17]. Furthermore binding phenotypes such as CD36 and EPCR binding appear to be primarily mediated by the tertiary structures of the CIDR domain and by hydrophobicity, characteristics that do not require an exact amino acid motif, thus allowing a high degree of sequence variation without affecting the function of the respective PfEMP1 domains [
18].
Considering the high variability of the
var gene family, it is remarkable that there is one
var gene that is conserved among all
P. falciparum strains. This gene,
var2csa, [
5] encodes the PfEMP1 VAR2CSA that binds to chondroitin sulfate A (CSA) in the human placenta and is primarily responsible for placental malaria during the first pregnancy [
6]. Because of this unique binding phenotype and function, there appears to be positive selection pressure to maintain this
var gene.
During a
var gene transcription analysis of a field isolate from Gabon (called MOA), a DBL with 100% sequence identity to the DBL of PF3D7_0617400 of the 3D7 reference genome [
19] was identified. In this work, the MOA allele of the gene (PFGA01_060022400) is shown to be 99% identical with PF3D7_0617400. The gene was detected in 36 of 714 African and Asian parasite suggesting a selective advantage for parasites carrying this
var gene.
Methods
Parasite lines and cultures
The MOA bulk culture was originally obtained from an chronic asymptomatic infection of a Gabonese individual as previously described [
19,
20]. A total of 19 clones were generated by limiting dilution in two independent cloning experiments [
19,
20]. The MOA C3 clone and the MOA D2 (PfGB01) [
21] clone were generated in the first cloning experiment directly after tissue culture adaptation of the MOA bulk culture [
19]. The NF54 A3 clone was generated from an NF54 bulk culture as described previously [
22]. Parasite stocks in glycerolite of the NF54 A3 [
22] laboratory strain, the culture adapted Gabonese MOA C3 [
19] and Δ MOA D2 (PFGB01) [
20] field isolates were thawed by slowly adding 5 drops of 12% NaCl solution and a 5 ml of 1.6% NaCl solution. The stocks were then spun down, the supernatant was discarded and the parasites were used to inoculate an in vitro culture.
The isolates 5259, 12295, 5420, 12480, 5798, 3256, 3324, 6022, 6210 were culture adapted from diagnostic specimens submitted for routine malaria diagnosis to the laboratory of the outpatient clinic of the Institute of Tropical Medicine in Tübingen, Germany. After conducting routine thick and thin blood smears, the remainder of the EDTA blood tube was centrifuged, the serum was separated and 500 µl erythrocyte pellet was used to inoculate 5 ml culture medium. After successful culture adaptation, parasites were expanded into a 20 ml culture and cryopreserved stocks as well as cell pellets were stored for future investigations. The Δ MOA D2 parasite line has been previously described [
20]. All
P. falciparum isolates were cultivated at 5% haematocrit of 0
+ erythrocytes from a local cell bank. RPMI 1640 medium was completed with 10% Albumax (Gibco), 25 mM HEPES Buffer, 2 mM
l-Glutamine and 0.05 mg/ml gentamicin (all PAA Laboratories). Parasites were incubated at 37 °C in 90% nitrogen, 5% oxygen and 5% carbon dioxide. Δ MOA D2 was cultured in medium without blasticidin as previously described [
20].
Cell pellets for DNA extraction were stored at − 20 °C. After thawing of the pellet, DNA was extracted using the QIAmp® DNA Blood Midi Kit (Quiagen, Cat. No. 51185) following the manufacturer’s protocol. DNA content was measured by Nanodrop® 1000 3.7.1 (Nanodrop Technologies).
Polymerase chain reaction (PCR) and agarose gel electrophoresis
For qualitative PF3D7_0617400 PCR, a specific primer set was designed, spanning exon 1 from promoter until the intron (Additional file
1: Table S1). Overlapping fragments for the entire exon 1 were generated. PCR with the
var2csa specific primers 10F and 75 R [
5] was employed to amplify an approximately 1700 bp fragment of
var2csa. The first 40 BP of exon 1 of PF3D7_0617400 were characterized with the UpsC promoter primer. This primer was originally developed by Rottmann et al. [
23] and has been previously evaluated on the MOA field isolate [
19].
5 µl genomic DNA were mixed with 31 µl H2O, 0.4 µl dNTPs, 3 µl MgCl, 5 µl buffer, 2.5 µl forward primer, 2.5 µl reverse primer, and 0.3 µl Taq-polymerase to a final volume of 50 µl per reaction. Standard PCR conditions were 94 °C for 3 min, 40× (94 °C for 10 s., 54 °C for 30 s., 72 °C for 30 s.), 72 °C for 3 min. For the primers pairs P3, P5 and P6 annealing and elongation temperatures were adjusted according to primer melting temperature and fragment length. Individual primer specific PCR conditions can be obtained from the author upon request. PCR fragments were separated using gel electrophoresis with a 1% agarose gel, 80–120 V current.
Sorbitol synchronization, RNA extraction and cDNA synthesis
20 ml parasite cultures were used for RNA extraction. The culture was pelleted, washed several times with 1× PBS and the erythrocytes were lysed with 0.02% saponin. The pellet was then washed 3 times with 1× PBS and resolved in 750 µl of Trizol® LS Reagent (Invitrogen). The samples were stored at − 20 °C until further processing.
For RNA extraction, 0.2 ml chloroform were added to 750 µl of thawed Trizol-lysate, shaken vigorously for 15 s and left at room temperature for 10 min. A 15 min centrifugation step at 12,000
g at room temperature followed for phase separation. RNA was extracted from the aqueous phase with the PureLink™ RNA Mini Kit (ambion by Life Technologies) according to the manufacturer’s protocol. RNA content was measured by Nanodrop
® 1000 3.7.1 (Nanodrop Technologies). All samples were treated with DNAse I
® (Invitrogen) according to the manufacturer´s protocol to remove any remaining DNA. cDNA was synthesized with random primers and Superscript II Reverse Transcriptase
® (Invitrogen) according to the manufacturer´s protocol. cDNA was tested for absence of DNA contamination by evaluation of proper splicing of the gene PFD1155w by PCR with the primer 5′GCAGGGAAAGGTTTTTCAAG3′ and the reverse primer 5′AAAGCTGAATCTTGGCCCGTT 3′ as described elsewhere [
22].
Var gene transcription analysis
cDNA DBL cloning
For analysis of specific
var gene expression in the MOAC3 clone, the active
var locus was identified by cloning cDNA
var PCR fragments obtained with universal primers [
24] followed by sequencing. The experiment was performed in two biological replicates directly after the original limiting dilution experiment of the MOA bulk culture [
19].
Quantitative real-time PCR
For quantitative RT-PCR reactions of the NF54 A3 strain, we employed the primer set of Salanti et al. [
5] with the modifications as previously described [
10,
19,
25] . For quantitative RT-PCR of the MOA field isolate the MOA primers C3_C3, D2_D2 and D5_D5 were used [
19]. All reactions included five housekeeping genes as controls: seryl-tRNA synthetase (PF3D7_0717700), fructose bisphosphate aldolase (PF3D7_1444800), actin (PF3D7_1246200), arginyl-tRNA synthetase (PF3D7_1218600) and glutaminyl-tRNA synthetase (PF3D7_1331700) [
25]. Reactions were performed at a final primer concentration of 0.25 µM using SensiMix SYBR No-ROX Kit (Bioline, QT650-05) in 20 µl reactions, measured in Corbett Research Rotorgene 3000 (95 °C for 3 min/95 °C for 15 s, 54 °C for 30 s, 68 °C for 30 s, 40 cycles/68 °C for 1 min). The same threshold was used for all analysis.
var gene copy numbers were determined relatively to PF3D7_1218600, using the ΔΔCT method [
5].
Targeted Sanger sequencing
For PCR fragment-sequencing, the PCR product was purified from an agarose gel or the PCR solution, using the Quiagen PCR purification kit according to the manufacturer’s protocol. Sequencing PCR was performed at a final reaction volume of 10 µl per reaction, containing 1µ BigDye, 2 µl reaction buffer, 2.5 µl forward or reverse primer, purified water and purified PCR product according to DNA concentration. Sequencing PCR conditions were: 94 °C for 10 s, 50 °C for 5 s, 60 °C for 4 min, back cycle to beginning: 25×. After PCR, all samples were cleaned up by Sephadex agarose column centrifugation. Sanger sequencing was performed in the Applied Biosystems ABI Prism 3130xl Genetic Analyzer.
Whole genome sequencing
Long-read Pacific Biosciences whole genome sequencing of the MOA D2 (PFGB01) and 5798 (PFTG01) parasite lines parasite was performed as previously described [
21].
Fragment analysis
57 microsatellites (MS) distributed over the 14 chromosomes of
P. falciparum were analysed by multiplex fragment analysis (Additional file
2: Table S2). DNA of all freshly culture adapted field isolates, the MOA C3 strain as well as the NF54 A3 were analysed. MS amplification and fragment length analysis were done as previously described [
26]. In four strains, approximately 30% of MS showed 2 amplification products/peaks (5798, 3324, 5420, 12480). The higher peak was arbitrarily chosen as the representative allele for the respective marker.
Calculation of expected heterozygosity
The expected heterozygosity (H
e) was calculated using the formula H
e = [n/(n − 1)][1 − Ʃpi2], where n is the total number of alleles at a distinct locus and reflects the proportion of the individual alleles as described previously [
27]. A new allele was defined as a difference of more > 3 bp between PCR fragments [
26]. The same criterion was applied for allele definition of the
var2csa fragments used for the H
e calculation of the
var2csa gene.
Assembling of sequences
Sequences were assembled and aligned using SeqScape software by Applied Biosystems.
Comparing sequences
The long read sequences were compared using the ACT software by Artemis [
28]. To compare the
var gene sequences with the
var gene database, a megablast (-e 1e-6 -m 8 -a 8 -v 15000 -b 15000 -F F) against the normalised data base from Otto et al. [
15] was done (varDB.Normalised.3 kb.nt.noVARx_noExon2.fasta.gz)–ftp
ftp://ftp.sanger.ac.uk/pub/project/pathogens/Plasmodium/falciparum/PF3K/varDB/NormalisedDataset/. To look for SNP around the conserved locus on chromosome 6, SNP were called with mpileup : (bcftools mpileup -r Pf3D7_06_v3 -f Pf3D7_06_v3.fasta $x.bam | bcftools call -cv -Ov –ploidy 1 -o SNP.var.$x.vcf).
Antibody preparation and Immunofluorescence assay (IFA)
The recombinant CIDRa2.1 domain of PF3D7_0617400 (MAL6P1.252) was produced at the Statens Seruminstitut, Cophenhagen, Denmark as described in [
18] and used to immunize mice. Blood of immunized mice was sent to the Institute of Tropical Medicine, Tübingen, Germany. Mouse sera containing IgG α PF3D7_0617400 mouse secondary antibody were depleted from RBCs by filtering through a MN 615 ¼ filter. IgG was purified using Protein G Spin Columns from Thermo Scientific according to the manufacturer’s protocol.
For IFA, very thin blood smears of parasite cultures were fixated in − 20 °C cold 100% methanol for 5 min and then stored at − 20 °C until further use. IFA was performed following a modified protocol as previously described [
29].
After a 5 min rehydration step in 1xPBS, slides were incubated for 1–2 h with anti CIDRa2.1 PF3D7_0617400 (diluted 1:50 in 1xPBS/1%BSA). The slides where then washed 3× with 1xPBS and incubated for 1 h with Alexa488 coupled mouse sera IgG α PF3D7_0617400 mouse secondary antibody (not diluted, 0.44 mg/ml). After another 3× washing with 1xPBS, slides were stained with Hoechst 33342 (diluted 1:1000) for 30 min. Slides were mounted over night with MOWIOL-488 and viewed through 100× oil immersion lens at a fluorescent microscope.
CD36 receptor binding selection
Human melanoma C32 (ATCC1, CRL-1585™) cells were used to select CD36 binding infected red blood cells (iRBCs) as previously described [
20].
Flow cytometry analysis
Flow cytometry analysis of MOA C3, NF54 A3 and Δ MOA D2 with serum of the MOA individual was performed as described in [
20].
Discussion
In this work, a conserved var gene PFGA01_060022400/PF3D7_0617400 (previously annotated as MAL6P1.252 /PFF0845c) is identified in parasites from Africa and Asia. PFGA01_060022400/ PF3D7_0617400 is located in the central cluster of chromosome 6. The predicted PfEMP1 domain structure of PFGA01_060022400/PF3D7_0617400 consists of a DBLa0.21, a CIDRa2.1, a DBLb4, a CIDRb1 and aDBLd1 domain. The only difference between the 3D7 and the field isolate allele is a small insertion of approximately 190 bp located between the DBLb4 and CIDRb1.
The CIDR a2.1 [
31] of PFGA01_060022400/PF3D7_0617400 possesses the recently described hydrophobic pocket that is responsible for PfEMP1 binding to the CD36 receptor [
18]. Consistent with this, NF54 parasites transcribing PF3D7_0617400 bound efficiently to CD36 receptors on human melanoma cells, showing that the expressed PfEMP1 exhibits the promiscuous CD36 binding phenotype. Although the hydrophobic pocket is present in virtually all CIDRa2-6, the sequence similarity is generally very low, as is the overall sequence similarity across the global population of CIDRa2-6 [
32]. This raises the question which binding phenotype might be conferred by the remainder of the PFGA01_060022400/ PF3D7_0617400 domains.
Metwally et al. [
31] conducted a comprehensive cytoadhesion analysis of the 3D7 laboratory strain on CHO-745 WT cells and in CHO-745 cells expressing recombinant CD36, Intracellular Adhesion Molecule 1 (ICAM) 1, P-selectin, E-selectin, CD9 and CD151. 3D7 parasites showed strong upregulation of PF3D7_0617400 transcription after selection on all cell types. The strongest upregulation was seen after binding selection on Chinese Hamster Ovary (CHO) wild type (WT) cells with PF3D7_0617400 being the only significantly upregulated
var gene (84% of the total
var gene signal). Together these data suggest that PF3D7_0617400 is able to bind to a yet unidentified receptor on CHO WT cells as well as CD36, ICAM, P-selectin, E-selectin, CD9 and CD15. Synergistic binding to multiple receptors has [
33] been shown to confer more efficient binding providing a potential explanation why PF3D7_0617400/ PFGA01_060022400 might confer a selective advantage.
The field isolate clone MOA C3 transcribed PFGA01_060022400 in a very stable fashion yet it was not possible to increase the low binding capacity of the strain to human melanoma cells expressing recombinant CD36. The MOA parasite line was originally obtained from a chronic asymptomatic infection that lasted for more than 90 days. The MOA parasites had therefore been under strong selective pressures prior to tissue culture adaptation. A previous phenotypic analysis of 19 clones from the MOA bulk culture showed that immune recognition of MOA clones did not correlate with
var gene transcription [
20]. Here, these observations are extended by phenotypic profiling of an NF54 and a MOA clone transcribing the same
var gene. CD36 selected NF54 A3 and MOA C3 parasites exhibited a marked difference in surface recognition signal and CD36 binding, yet the corresponding PfEMP1 was detected by immunofluorescence assay (IFA) in both cell lines. This suggests a difference in PfEMP1 display between the two cell lines. Several investigations have reported that PfEMP1 expression is influenced by semi-immunity [
20,
34‐
36] and Hoo et al. [
37] have shown recently that the duration of replication in the human host has a strong impact on the
P. falciparum transcriptome. It is thus tempting to hypothesize that the prolonged intra-host replication of the MOA parasites selected for parasites with reduced PfEMP1 display to allow the submicroscopic parasitaemia of chronic infections. It is conceivable that expression of a PfEMP1 variant with synergistic binding to multiple receptors provides a selective advantage during chronic infections. Future investigations with culture adapted parasites from semi-immune individuals are necessary to further address this question.
PFGA01_060022400 was identified in 37 isolates of a recently characterized global population of 714 fully sequenced isolates. A previous analysis of this population had identified several var genes with fragments > 3.5 kb that were shared at 99% percent identity in up to 55 isolates. These genes are located on chromosomes 7, 4, and 8 in the areas of drug resistance associated selective sweeps. An additional conserved var gene was located in a telomeric genetic sweep of chromosome 6. In contrast the parasite isolates carrying PFGA01_060022400/ PF3D7_0617400 showed no evidence of a genetic sweep in chromosomal areas flanking the conserved locus in the central cluster of chromosome 6. This suggests that PFGA01_060022400/PF3D7_0617400 might provide some independent “fitness advantage”. The population analysis also showed that the DBLb4 and CIDRb1 sequences were present in more isolates than the DBLa0.21, CIDRa2.1, DBLb4, CIDRb1 sequences, indicating that fragments of the gene are conserved to different degrees in the global P. falciparum population, suggesting that individual sequence blocs might be under “positive selection”.
To further evaluate this hypothesis, the genetic diversity of a highly variable and highly conserved part of the DBL2x domain of the
var2csa gene and the genetic diversity of 57 MS [
26] were compared in the original population of 10 field isolates. As expected the genetic diversity of MS was high (0.76–0.8) and in the range of previously reported values for MS genetic diversity [
27,
38‐
40]. Strikingly, the variable DBL2x part exhibited the same average genetic diversity (H
e = 0.8) as the MS, suggesting that sequence variation in these parts of the gene have no deleterious effect on parasite fitness. In contrast the genetic diversity of the conserved DBL2x fragment was 0 and indeed the sequence was completely identical across all isolates. These data are consistent with a recent analysis of the
var2csa gene in a global population of >2000 isolates that showed a high sequence diversity along the DBL2x but a high degree of conservation in the area corresponding to the conserved fragment [
41].
While
var2csa is present in every parasite, the full length PFGA01_060022400/ PF3D7_0617400 was only detected in approximately 5% of field isolates. PFGA01_060022400 /PF3D7_0617400 is located in the central cluster of chromosome 6 and belongs to the UpsC subclass of
var genes. UpsC
var genes have been shown to be preferentially transcribed during long term in vitro culture and during chronic asymptomatic infections [
19,
22,
42]. The binding phenotype of PFGA01_060022400 /PF3D7_0617400 may, therefore, be important to establish chronic asymptomatic infections. If receptor binding is indeed responsible for the conservation of PFGA01_060022400 /PF3D7_0617400 this would imply that other
var genes must confer the same binding phenotype. The recent global analysis of the
var gene family by Otto et al. [
15] showed that UpsC genes are the most conserved among the
var gene classes which could be consistent with a common binding phenotype mediated by UpsC
var genes. In summary, the data reported here suggests that individual PfEMP1 binding phenotypes may limit the sequence diversity of individual members of the
var gene family.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.