The Duffy binding protein (PkDBPαII) of Plasmodium knowlesi from Peninsular Malaysia and Malaysian Borneo show different binding activity level to human erythrocytes

Malaria is a parasitic disease that affects millions of people in Africa, Asia, and South and Central America. In 2015, there were approximately 212 million new cases of malaria, causing nearly 429,000 deaths [1]. This mosquito-borne disease is caused by blood protozoa of the genus Plasmodium. Four species, Plasmodium falciparum, Plasmodium vivax, Plasmodium malariae, and Plasmodium ovale (now comprises of P. ovale wallikeri and P. ovale curtisi) have long been known to cause human malaria [2, 3]. Over the past decade, the simian malaria parasite Plasmodium knowlesi has emerged to cause significant human infections in Southeast Asia [4‐8]. Microscopically, P. knowlesi bears a resemblance to the benign P. malariae, but it has a unique 24-h erythrocytic cycle (quotidian fever pattern), and it can cause severe disease and death if left untreated. Knowlesi malaria is now the most common form of malaria in Malaysia especially in Malaysian Borneo [9‐11].

Plasmodium merozoites invasion into erythrocyte is a complex process which involves attachment, apical reorientation, tight-junction formation and entry into a parasitophorous vacuole. These steps are mediated by specific molecular interactions between the parasite’s ligand and its corresponding receptor on the surface of erythrocyte membrane [12]. Both P. knowlesi and P. vivax are known to interact with the Duffy antigen receptor for chemokines (DARC) to invade Duffy-positive human erythrocytes. Duffy-negative human erythrocytes are refractory to invasion by these two Plasmodium species [13‐15]. The invasion of human erythrocytes by P. knowlesi is dependent on the interaction of DARC with the parasite’s ligand, the Duffy binding protein (PkDBPα) [16]. PkDBPα can be divided into seven regions (I–VII). The N-terminal cysteine rich region II (PkDBPαII) contains the critical Duffy-binding-like (DBL) ligand domain for binding to the erythrocyte [17, 18]. PkDBPαII has been shown to bind to Duffy-positive human and macaque erythrocytes [19]. On the other hand, two other related proteins, the PkβII and PkγII, bind only to macaque erythrocytes [17].

The clinical symptoms of malaria are primarily attributed to the blood-stage of the parasite life cycle, which results from repeated rounds of erythrocyte invasion, parasite multiplication, erythrocyte lysis and release of free merozoites. It has been observed that antibodies raised against PkDBPαII could inhibit P. knowlesi invasion of human and macaque erythrocytes in vitro [20], making it a possible target vaccine candidate against knowlesi malaria.

Although knowlesi malaria is seen in both Peninsular Malaysia and Malaysian Borneo, hyperparasitaemia and severe cases are more prominent in Malaysian Borneo [10, 21‐27]. Previous studies have shown genetic diversity in the PkDBPαII of clinical isolates from Peninsular Malaysia and Malaysian Borneo [28, 29]. It was further observed that PkDBPαII haplotypes from Peninsular Malaysia and Malaysian Borneo were genetically distinct. This led us to investigate whether the genotypic differences in PkDBPαII could influence the ability of the parasite for invasion into erythrocytes. In the present study, binding activity of Peninsular Malaysia and Malaysian Borneo PkDBPαII haplotypes with human and macaque (Macaca fascicularis) erythrocytes was compared using the erythrocyte-binding assay.

The PkDBPαII of P. knowlesi isolates from Peninsular Malaysia and Malaysian Borneo were successfully transfected and expressed on COS-7 cells (Fig. 1). These transfected COS-7 cells were tested with erythrocytes in the erythrocyte-binding assays. In the negative controls which consisted of non-transfected COS-7 and COS-7 cells transfected with plasmid without inserts, did not produce any fluorescence.

Recombinant PkDBPαII expressed on the COS-7 cell surface was able to bind to both human and macaque erythrocytes, but at different levels (Fig. 2). In the erythrocyte-binding assay using human erythrocytes, the binding activity or number of rosettes formed with PkDBPαII of Peninsular Malaysia was three times higher than that of Malaysian Borneo PkDBPαII (156.89 ± 6.62 rosettes and 46.00 ± 3.57 rosettes, respectively; P < 0.0001) (Table 1). When assayed with macaque erythrocytes, the PkDBPαII of Peninsular Malaysia (356.56 ± 6.75 rosettes) and Malaysian Borneo (355.22 ± 11.69 rosettes) showed similar binding activity levels (Table 1).

A change in binding activity was observed when the human erythrocytes were treated with anti-Fy6 prior to erythrocyte-binding assay. The binding activity with PkDBPαII of Peninsular Malaysia and Malaysian Borneo was completely eliminated (Fig. 3a–d). However, anti-Fy6 treatment of macaque erythrocytes had no effect on the binding activity of PkDBPαII (Fig. 3e–h). The PkDBPαII expressed on the surface of COS-7 cells bound to both non-treated and anti-Fy6-treated macaque erythrocytes regardless of the expressed PkDBPαII haplotype. In fact, the binding activity using macaque erythrocytes were significantly higher than those incubated with human erythrocytes (Table 1). Furthermore, the intensity of rosette formation in the PkDBPαII of Peninsular Malaysia and Malaysian Borneo was different when incubated with human and macaque erythrocytes. The rosettes formed on COS-7 cells with macaque erythrocytes were generally larger when compared to the rosettes formed with human erythrocytes irrespective of the PkDBPαII haplotypes (Fig. 3).

H2 and H47 are the predominant PkDBPαII haplotypes in Peninsular Malaysia and Malaysian Borneo, respectively. These haplotypes were chosen for this study, with the aim of determining whether haplotypic differences would affect PkDBPαII binding activity with erythrocytes.

Results from this study revealed higher binding activity of PkDBPαII of Peninsular Malaysia with human Duffy-positive erythrocytes, as compared to Malaysian Borneo PkDBPαII. The PkDBPαII expressed on the COS-7 cell surface was based on the gene region defined from previous studies [30, 31]. This region encompasses a subdomain that provides the essential amino acid residues for proper folding and binding with the DARC on human erythrocytes. Within this subdomain, there are 12 cysteine residues (at positions 16, 29, 36, 45, 99, 176, 214, 226, 231, 235, 304, 306) that form 6 disulphide bridges which contribute to the tertiary structure of PkDBPαII for interaction with DARC. These cysteine residues have been shown to be conserved in the PkDBPαII of Peninsular Malaysia and Malaysian Borneo [28]. Amino acid residues Y94, N95, K96, R103, L168 and I175 in this subdomain have been identified as essential for binding with DARC (Fig. 4). Sequence alignment revealed that these residues were conserved except at position 95, where N was substituted with D in the PkDBPαII of Malaysian Borneo (Fig. 4). Apart from this N95D substitution, 11 other amino acid differences were identified between the PkDBPαII of Peninsular Malaysia and Malaysian Borneo (Fig. 4). These amino acid substitutions may have caused changes in the conformational structure of PkDBPαII of Malaysian Borneo, thus affecting its ability to bind to human erythrocytes efficiently.

Surprisingly, the binding activity level of PkDBPαII of Peninsular Malaysia and Malaysian Borneo to macaque erythrocytes was similar. The reason for this is unclear, but it may be plausible that amino acids for interaction with macaque erythrocyte receptor may be different from those of human erythrocyte. Therefore, the non-synonymous mutations in the PkDBPαII of either Peninsular Malaysia or Malaysian Borneo have no effect on the binding level of the protein to macaque erythrocytes.

The PkDBPαII plays a vital role in the invasion of parasites by mediating interaction with its corresponding receptor, the Duffy antigens on the surface of human erythrocytes. Previous studies have shown that human Duffy-negative Fy(a−b−) erythrocytes are refractory to invasion by P. knowlesi and P. vivax merozoites, while human Duffy-positive erythrocytes are susceptible [32, 33]. Plasmodium knowlesi merozoites require the interaction between PkDBPαII with Duffy determinant Fy^a or Fy^b for binding with erythrocytes. Thus, by blocking these determinants, it would inhibit the binding. This inhibition was observed in the present erythrocyte-binding assay using anti-Fy6 treatment. The Fy6 epitope is located at the N-terminal extracellular domain of DARC in close proximity to the Fy^a/Fy^b determinant [34]. A previous study has shown that incubation with anti-Fy6 resulted in full inhibition of binding activity and invasion by P. knowlesi into human erythrocytes [35]. The inability of PkDBPαII of Peninsular Malaysia and Malaysian Borneo to bind to anti-Fy6 treated human erythrocytes demonstrated that the expressed PkDBPαII bound to the Duffy determinants of human DARC and not to other available receptors, such as sialic acid residues [36, 37].

In the present study, treatment of macaque erythrocyte with anti-Fy6 did not affect its binding with PkDBPαII. The Fy6 determinant is present on all human erythrocytes except those with Fy(a−b−) phenotype. In contrast, the Fy6 determinant is present only on the erythrocytes of some non-human primate species such as the Great Apes and New World monkeys. Erythrocytes of M. fascicularis are Fy^b but lack the Fy6 determinant [38]. Hence, the inability of anti-Fy6 to hinder interaction of PkDBPαII with M. fascicularis erythrocytes is due the absence of the Fy6 determinant. With the Fy^b determinant on erythrocytes still exposed, the binding of PkDBPαII with treated M. fascicularis erythrocytes occurred successfully.

The present study also demonstrated preferential binding activity of PkDBPαII. There was lesser number of rosettes in the erythrocyte-binding assay with human erythrocytes compared to macaque erythrocytes. This may suggest that the PkDBPαII binds to different Duffy determinants on human and macaque erythrocytes. This finding is in agreement with a previous study that postulated alternative determinants on DARC for binding [38].

The human Duffy-positive Fy(a+b−) erythrocytes were chosen in the present study primarily because this phenotype is the most common in the Malaysian population. Nonetheless, it would be worthwhile in future studies to investigate other Duffy phenotypes such as Fy(a+b+), Fy(a−b+) and Fy(a−b−). In vitro invasion studies have shown reduced efficiency of P. knowlesi to invade human Fy(a+b−) compared with Fy(a−b+) erythrocytes [14]. Furthermore, Fy^a expressing erythrocytes show reduced binding to PvDBPII and reduced susceptibility to vivax malaria [39].

There has been postulation that enhanced virulence and multiplication rates of Plasmodium parasites may be results of genetic polymorphisms that improve binding ability of the parasites to human erythrocytes [40]. Clinical and epidemiological findings revealed high number of severe knowlesi malaria in Malaysian Borneo. Oddly, results from this study found lower binding activity level of the Malaysian Borneo PkDBPαII haplotype compared to that of Peninsular Malaysia. A study has reported that P. knowlesi normocyte binding protein (PkNBPXa) polymorphisms are important determinants of high parasitaemia and disease severity in P. knowlesi infection [41]. Recently, it has been demonstrated that PkNBPXa is required for human erythrocyte invasion, and suggested that its role occurs after initial merozoite attachment [40]. Nevertheless, further research needs to be conducted to determine whether there is interaction between PkDBPαII and PkNBPXa in the invasion process, and polymorphism of PkDBPαII in part contributes to disease severity.

This study is the first report of phenotypic difference between PkDBPαII haplotypes. The biological implication of this finding is yet to be determined. Therefore, further studies need to be carried out to determine whether this differential binding level can be associated with severity of knowlesi malaria in human.