Limited polymorphism in Plasmodium falciparum ookinete surface antigen, von Willebrand factor A domain-related protein from clinical isolates

Background As malaria becomes increasingly drug resistant and more costly to treat, there is increasing urgency to develop effective vaccines. In comparison to other stages of the malaria lifecycle, sexual stage antigens are under less immune selection pressure and hence are likely to have limited antigenic diversity. Methods Clinical isolates from a wide range of geographical regions were collected. Direct sequencing of PCR products was then used to determine the extent of polymorphisms for the novel Plasmodium falciparum sexual stage antigen von Willebrand Factor A domain-related Protein (PfWARP). These isolates were also used to confirm the extent of diversity of sexual stage antigen Pfs28. Results PfWARP was shown to have non-synonymous substitutions at 3 positions and Pfs28 was confirmed to have a single non-synonymous substitution as previously described. Conclusion This study demonstrates the limited antigenic diversity of two prospective P. falciparum sexual stage antigens, PfWARP and Pfs28. This provides further encouragement for the proceeding with vaccine trials based on these antigens.


Background
Development of effective Plasmodium falciparum vaccines is a global health priority with all stages of the malaria lifecycle being investigated for possible vaccine strategies [1]. Mosquito-stage vaccines aim to induce immunity to the forms of the parasite found in the mid-gut of mosquitoes: the sexual stage macro-and microgametes and the postfertilization zygotes and ookinetes. Antibodies that target antigens in these stages prevent infection in the mosquito thus breaking the transmission cycle. Since malaria trans-mission is both highly localized and focal [2], such vaccines may form a useful element as part of integrated control programmes to locally eliminate or substantially reduce transmission. They may also have a useful role in conjunction with chemotherapeutic agents to prevent drug-resistant or vaccine-escape mutants.
The P. falciparum, mosquito-stage vaccine candidate antigens include Pfs25, Pfs28, Pfs48/45, Pfs230 and the recently described, von Willebrand factor A domain-related protein (PfWARP) [3,4]. PfWARP (PF08_0136b) is located on chromosome 8 (GenBank: NC_004329 Region: 1221207...1222079) and encodes a 290 amino acid protein. On the basis of previous study of orthologues in Plasmodium berghei and Plasmodium gallinaceum, PfWARP is thought to be a soluble micronemal protein expressed in the late ookinete and early oocyst [5][6][7]. It is proposed that the von Willebrand factor A domains found in PfWARP and other malaria antigens, including circumsporozoite protein and thrombospondin-related anonymous protein have a role as an adhesive substrate and may assist in host cell invasion [7,8].
Pfs28 (PF10_0302) is a lead candidate for a transmissionblocking vaccine. It is a 28-kDa protein and is presumably anchored to parasite surface by glycosylphosphatidylinositol [9]. It is encoded on chromosome 10 and shares putative functional determinants with Pfs25. Pfs28 is expressed on the surface of late ookinetes of P. falciparum and it is proposed that it is involved in adherence to the mosquito's gut epithelium [10]. Previous studies have examined Pfs28 polymorphism in laboratory isolates Dd2, 2D11, CAMP, LE5, LF4, HB3, 7G8 [11] and from 32 field isolates from a geographically restricted region in the Philippines [12]. These have shown the presence of a single non-synonymous A274G base substitution, resulting in a conservative amino acid change from lysine to arginine (GenBank: L25843).
As part of the assessment of PfWARP's suitability as a vaccine candidate, the diversity in this gene was assessed using P. falciparum field isolates from a range of geo-graphic locations. In addition, Pfs28 was sequenced in the same isolates to extend previous observations of its' diversity.

Methods
Review and approval of the study was obtained from the Research Ethics Committee of the Royal Brisbane Hospital and the Melbourne Health Human Research Ethics Committee. Clinical samples from P. falciparum-infected patients were obtained from the Malaria Reference Laboratory at the Royal Brisbane Hospital, Queensland. Twenty-two samples were studied from unrelated travellers returning from Papua New Guinea, Solomon Islands, Pakistan, Kenya, Tanzania, Uganda, Zimbabwe, and Ghana.
Whole blood specimens were stored in 8 M guanidine hydrochloride/0.1 M Na H 2 PO 4 . Genomic DNA was isolated from the samples using the QIAquick ® PCR purification kit (Qiagen, Hilden, Germany). PfWARP and Pfs28 polymorphism was assessed using PCR conditions as outlined (Table 1). All PCR reactions used Amplitaq Gold ® DNA polymerase (Roche Diagnostics, Mannheim, Germany). PCR products were detected after submarine electrophoresis using 1% agarose gel stained with ethidium bromide. PCR products were purified for direct sequencing using QIAquick ® PCR purification kit (Qiagen, Hilden, Germany) or were agarose gel purified using MinElute Gel extraction kit ® (Qiagen, Hilden, Germany) purification kit. DNA sequencing was performed utilising the automated 3100 Genetic Analyser capillary system. In addition, samples with multiple infections of P. falciparum were detected by merozoite surface protein 2 (MSP2) PCR as previously described [13]. Amplification of this highly polymorphic gene allowed rapid detection of size separable fragments by gel electrophoresis.

Results
PfWARP was amplified from 19 of the field samples and three laboratory strains (3D7, FC27, and FCR3). As judged by MSP2 genotyping, there were multiple infections in 7 of the 19 patient samples. These 7 patient samples had a minimum of 26 genetically different clones of P. falciparum. However, direct sequencing of PCR products from each of these samples showed only a single genotype of PfWARP with no mixed peaks in the sequencing traces. Non-synonymous nucleotide substitutions were found to occur in three positions amongst the variant sequences.
No synonymous substitutions were identified. The predicted amino acid changes were conservative and contained within the single Von Willebrand factor domain ( Table 2). Sample numbers were insufficient to determine whether there was any geographic distribution to the PfWARP alleles.
Pfs28 was amplified in 22 field samples (representing at least 31 clones due to the existence of multiple MSP2 alleles in nine patients) and in 3D7. The A274G, non-synonymous, nucleotide substitution as previously described, was found with 11 of the 22 samples having the A274 allele (Table 2). No other nucleotide substitutions were present. This data was combined with that from Hafalla et al [12] to consider the geographic distribution of Pfs28 alleles. In Asia/Pacific isolates, 42 of 48 carried the G274 allele while only 1 of 14 of Africa/Americas isolates bore this allele (p < 0.001, Fisher's exact test). While recognizing the size limitations of this study, it is still striking that there is such a significant difference in the geographic distribution of these Pfs28 alleles (Table 2).

Discussion
For the first time, this study addresses the issue of antigenic diversity in PfWARP. This antigen appears to have restricted diversity with infrequent point mutations as defined. Additionally, by documenting the genotype of Pfs28 in field isolates from a larger cohort with wider geographic distribution than in previous studies, we confirm that there are only two alleles of this protein. Restricted antigenic diversity has been described in other sexual stage antigens and is explicable due to the absence of human immune selection pressure [14][15][16]. By contrast, asexual stage candidate antigens such as MSP2 are highly polymorphic due to P. falciparum's immune evasion mechanisms [17].
Pfs28, and its orthologues from other species, Pvs28 (Plasmodium vivax) and Pbs21 (P. berghei) clearly induce anti-bodies that block parasitic infection in mosquitoes and are thus candidate antigens for inducing transmission blocking immunity in humans either alone [18] or in combination with other mosquito-stage antigens. So far there have been two published vaccine trials in humans with Pfs25 [19] and with its P. vivax orthologue, Pvs25 [20], with other trials currently underway.

Conclusion
The demonstration of limited sequence variation in the newly described P. falciparum ookinete surface antigen PfWARP and the confirmed highly conserved nature of Pfs28 in clinical samples from a wide geographical distribution, increase the confidence that these proteins may also be suited for mosquito-stage vaccine development.

Authors' contributions
JSR carried out the genetic studies, sequence alignment and drafted the manuscript. NJM designed and supplied the primers. DPE designed a study of returned travellers with malaria and recruited the patients, assisted in the sequence alignment and helped draft the manuscript. All authors read and approved the final manuscript.