Molecular and cellular interactions defining the tropism of Plasmodium vivax for reticulocytes

Abstract

Plasmodium vivax is uniquely restricted to invading reticulocytes, the youngest of red blood cells. Parasite invasion relies on the sequential deployment of multiple parasite invasion ligands. Correct targeting of the host reticulocyte is mediated by two families of invasion ligands: the reticulocyte binding proteins (RBPs) and erythrocyte binding proteins (EBPs). The Duffy receptor has long been established as a key determinant for P. vivax invasion. However, recently, the RBP protein PvRBP2b has been shown to bind to transferrin receptor, which is expressed on reticulocytes but lost on normocytes, implicating the ligand-receptor in the reticulocyte tropism of P. vivax. Furthermore there is increasing evidence for P. vivax growth and sexual development in reticulocyte-enriched tissues such as the bone marrow.

Introduction

Malaria, caused by Plasmodium parasites, is a major public health burden with estimations of 200 million cases and over 400,000 deaths in 2015 [1]. The species P. vivax, accounts for the majority of infections outside the African continent, causing over 13 million cases of malaria in 2015. Upon infection of a new host, the parasite undergoes one round of replication in the liver followed by establishment of a cyclical infection in red blood cells (RBCs), leading to all of the clinical symptoms of disease [2]. Invasion of new RBCs by free merozoites occurs rapidly and relies upon the interactions between parasite-expressed invasion ligands and target host receptors [3]. P. vivax is unique among human malaria parasites in being strictly restricted to invading the youngest red blood cells known as reticulocytes. Reticulocytes make up only 1 – 2% of the circulating RBCs, thus indicating free merozoites accurately identify this subset of host cells during invasion. However, our understanding of the process of parasite invasion remains deficient, in part due to the lack of a continuous in vitro culture system for P. vivax. This is critical as many invasion ligand proteins are being considered as vaccine candidates [4]. In this review we summarize the latest advances in our understanding of P. vivax invasion of reticulocytes.

Parasite invasion.

The majority of our mechanistic understanding of Plasmodium spp. invasion of RBCs comes from studies of other human (P. falciparum, P. knowlesi) and rodent (P. berghei, P. yoelii) parasites due in part to our limited ability to culture P. vivax. Invasion of RBCs is a rapid process (< 1 min), but during that time the newly released merozoite undergoes a highly orchestrated series of events that lead to recognition, binding and entry of a new host cell [3] (Figure 1a). Invasion machinery is deployed at the merozoite surface (where initial contact with the host cell takes place) and in specialized organelles, micronemes and rhoptries. Invasion ligands are released from these organelles to enable strong attachment to the host cell and subsequent tight junction formation. The tight junction is formed between two parasite-produced proteins: AMA1 on the merozoite surface, and the rhoptry-resident RON2 complex which is injected into the host cell membrane. Parasitophorous vacoule formation follows parasite entry via the action of the moving junction complex. The ability of the parasite to identify the correct host cell relies on interactions between parasite adhesins and surface membrane proteins during the initial steps of invasion. It is assumed that there is conservation in the steps of the invasion process, however this remains to be formally demonstrated. There have also been a limited number of global gene expression [5–7] and proteomic studies [8–10] of P. vivax, however our knowledge of invasion ligand usage during invasion remains sparse. In addition, the identities of the host receptors used during invasion have not been fully elucidated, and in particular there is a lack of knowledge of the reticulocyte tropic host factors of P. vivax.

Figure 1.

P. vivax invasion of host cells. (A) Schematic showing the different steps during merozoite invasion of host RBC from initial weak attachment, reorientation, tight junction formation and invasion. (B) List of known and hypothesized P. vivax parasite invasion ligands along with summary reports of RBC binding, either directly or from orthologues, localization data to merozoite surface, rhoptry or microneme organelles and reactivity with patient sera.

Parasite adhesins – multigene family proteins.

The exterior of the merozoite is decorated by a large number of merozoite surface proteins (MSPs) which are thought to mediate the initial interactions with the host cell [3] (Figure 1b). The MSPs have been best studied in P. falciparum where PfMSP1 forms a number of multiprotein complexes with other MSPs including PfMSP3, PfMSP6, PfMSP7, PfMSP9, PfMSPDBL1 and PfMSPDBL2 [11]. PfMSP1 has also been shown to interact with the two most abundant RBC membrane proteins: band 3 and glycophorin A [12]. P. vivax also has both an MSP1 orthologue (PvMSP1) and a MSP1 paralogue (PvMSP1P) which colocalizes with PvMSP1 and may have a similar function [13]. However, there is little known about the presence of analogous PvMSP complexes, and P. vivax only has orthologues of two complex-forming MSPs: the 11-member multigene PvMSP7 (PfMSP7 orthologues) and PvMSP9 (PfMSP9 orthologue) [14]. The PvMSP3 multigene family is unrelated to the P. falciparum MSP3 [15]. A number of the PvMSP3 proteins have been biochemically characterized and are mostly localized to the merozoite surface, except PvMSP3.7, which has apical localization [16]. Finally, PvMSP4 and PvMSP10 have P. falciparum orthologues that show RBC binding [14].

PvDBP and DARC:

Two invasion ligand gene families, the erythrocyte binding proteins (EBPs) and reticulocyte binding proteins (RBPs) have been extensively studied in P. falciparum [3]. EBP and RBP proteins bind to various RBC host receptors and play an important role in P. falciparum host tropism. In addition there is functional redundancy with many of the invasion ligands, giving the parasite multiple complementary invasion pathways. However, much less is known about the existence of discrete invasion pathways in P. vivax.

P. vivax has two EBP family members: P. vivax Duffy-Binding Protein (PvDBP) [17] and P. vivax Erythrocyte Binding Protein (PvEBP) [18] (Figure 1b). PvDBP binds to the Duffy Antigen Receptor for Chemokines (DARC/Duffy) [19]. DARC is polymorphic and a variant that results in no DARC expression on RBCs (Duffy negative) is highly prevalent in sub-Saharan Africa and is thought to have been selected for by P. vivax. DARC also has a high frequency polymorphism at position D42 (FyA polymorphism) which results in reduced PvDBP binding [20]. The PvDBP/DARC interaction is almost universally required for P. vivax infection. Troublingly, there have been increasing reports of P. vivax found in DARC-negative individuals [21], suggesting that the parasite is evolving to overcome this invasion blockade which could have serious implications for the spread of malaria. One proposed mechanism for DARC-negative invasion is that duplications of PvDBP, as has been reported in multiple studies [22–24], may enable the parasite to utilize an alternative host receptor. Alternatively, one of the PvRBP proteins may functionally compensate for PvDBP in DARC-negative invasion [24]. However further studies are needed in order to understand the mechanism of DARC-negative invasion.

The second EBP member, PvEBP was recently described during sequencing of a clinical P. vivax isolate [18]. Recombinant PvEBP has been reported to bind to DARC-positive reticulocytes, but not DARC-negative reticulocytes, suggesting a role as an alternative adhesin [25].

Reticulocyte binding proteins.

The tropism of P. vivax for reticulocytes is thought to be mediated by the seven full-length RBPs [14] (Figure 1b). The RBP orthologues in P. falciparum, the reticulocyte homolog (Rh) proteins, play critical roles during invasion, but do not target P. falciparum to reticulocytes [3]. The interaction of PfRh5 with the host protein basigin was found to be essential for P. falciparum invasion [26]. While there is no direct orthologue of PfRh5 in P. vivax, other accessory proteins that bind PfRh5 (CyRPA, RIPR, P113) [27–29] are all present in P. vivax, suggesting that they may bind an alternative protein during invasion.

Recently, there has been an improved understanding of the PvRBPs, including studies of genetic diversity [30], gene expression in clinical isolates [31] and immune reactivity with patient sera [32]. Most PvRBPs are very large proteins, often exceeding 2000 amino acids in length, making it challenging to express full-length proteins. Recent studies have succeeded in expressing sub-domains of PvRBPs that are capable of binding to RBCs, and indeed these studies have demonstrated reticulocyte tropism for PvRBP1a [33–35], PvRBP1b [33], PvRBP2b [31,32], PvRBP2c [34] and PvRBP2P2 [32]. PvRBP2a shows binding to both reticulocytes and normocytes [36], suggesting that it may not interact with a solely reticulocyte tropic host receptor.

TfR is a host receptor for P. vivax.

Initial studies of recombinant PvRBP2b protein showed very high levels of binding to reticulocytes [32], suggesting that it may bind an abundant reticulocyte specific receptor. Subsequent biochemical evidence demonstrated direct interaction between purified PvRBP2b and the apical domain of transferrin receptor (TfR/CD71) [37]. TfR is an essential gene abundant in developing erythroid cells and young reticulocytes and is responsible for binding and internalization of iron-loaded transferrin. The PvRBP2b/TfR interaction was functionally validated by generating a TfR mutant using a novel JK-1 erythroleukemia cell line [38], and this mutant resulted in abolished PvRBP2b binding and substantially reduced P. vivax invasion, indicating that TfR is likely an essential receptor for P. vivax. Interestingly, TfR also functions as a receptor for multiple arenaviruses [39], and there is overlap in the binding site between Machupo virus GP1 and PvRBP2b, possibly due to the accessibility of this domain. However, there is little known about TfR polymorphisms and whether any mutations can reduce PvRBP2b binding and P. vivax invasion as seen for mutations in DARC. Furthermore, the likely essentiality of the TfR/PvRBP2b interaction makes this system an attractive candidate for vaccine development and for immuno-epidemiology studies for protective host-antibodies [40].

Other invasion adhesins.

P. vivax has orthologues to a large range of candidate P. falciparum adhesins (Figure 1b). P. vivax has an expanded family of tryptophan rich antigen proteins, the TRAgs (Pv-fam-a), with 36 identified members [14]. Ten of these proteins have been shown to bind to RBCs and based on gene expression studies, at least four of the ten TRAgs are likely involved in rosette formation, four are likely to be involved in RBC invasion and the remaining two do not have expression data [41]. Recently it was shown that PvTRAg38 is capable of binding Band 3 [42] and basigin [43] although it is unclear at what stage of invasion these interactions might function. PvTRAg56.6 does not have direct RBC-binding functions, however, it does bind to PvMSP7, and it has been suggested that the pair can form a complex with PvMSP1 [44].

The surface expressed 6-Cys family of proteins are involved in protein interactions and are important in several stages of the parasite life cycle including exflagellation of male gametocytes, evasion of the mosquito immune system, parasitophorous vacuole formation, complement lysis protection and parasite invasion [45]. Two members, PvP12 and PvP41, form a complex on the merozoite surface and interact with another uncharacterized protein, PVX_110945 [46]. The P. falciparum orthologue of PvP38 binds RBCs, while the P. falciparum orthologue of PvP92 is an important inhibitor of complement-dependent lysis of the free merozoite [45].

There are also other candidate adhesins which remain poorly characterized and include: surface antigens PvAARP [47], PvRBSA [48], PvMSA180 [49]; the rhoptry antigens Pv34 [46], PvRAMA [50], PvRhopH3 [51], PvRALP1 [52] and PvRA [53]; and the micronemal antigens PvGAMA [54,55], PvMTRAP [46] and PvMA [46]. In addition there are a number of uncharacterized candidate adhesins identified in a protein expression screen [46], for which little functional data is currently available.

There is an emerging understanding of the functions of the multiple invasion ligands during P. vivax invasion. Recently there have been several efforts to produce libraries of recombinant Plasmodium invasion ligands and these have been screened using patient sera in order to identify targets of naturally-acquired immunity [46,56–59]. In one study, a combination of five antigens (PvDBP, PvEBP, PvRBP1a, PvCyRPA and PVX_081550) were associated with a reduced risk of P. vivax [57]. It will be critical to extend these types of studies with functional characterization with the goal of identifying conserved, essential invasion ligand-host receptors for prioritization as vaccine candidates.

Reticulocyte heterogeneity and P. vivax.

During erythroid differentiation, bone marrow CD34+ hematopoietic stem cells undergo programmed differentiation through a series of distinct stages culminating in orthochromatic cells that undergo nuclear loss to form a pyrenocytes and a nascent reticulocyte (Figure 2). Reticulocytes themselves undergo further maturation over the course of 48–72 hours, split evenly between the bone marrow and then the peripheral circulation where the mature RBC is formed [60]. Maturing reticulocytes undergo substantial cellular changes including loss of sub-cellular organelles, membrane remodeling via exosome formation and changes to the cytoskeleton resulting in increased deformability. Recent studies have highlighted some of the metabolic [61] and proteomic changes that accompany this process [62,63].

Figure 2.

Reticulocyte maturation and susceptibility to parasite invasion. CD34+ hematopoietic stem cells (HSCs) mature in the bone marrow lumen to form orthochromatic erythroblasts. These cells undergo a process of enucleation, producing a nascent reticulocyte and a pyrenocyte. Reticulocytes continue to mature initially in the bone marrow and subsequently in the peripheral circulation, eventually forming mature RBCs. Schematic levels of transferrin receptor (TfR) and susceptibility of cells to P. vivax and P. falciparum is indicated.

Reticulocyte heterogeneity has an impact on the ability of P. vivax parasites to invade erythroid cells. In ex vivo assays, Malleret et al. observed higher rates of parasite invasion into young reticulocytes, which have high levels of TfR [64]. Indeed, a subsequent study identified TfR as a host receptor for P. vivax invasion [37]. Throughout reticulocyte maturation, TfR is incorporated into exosomes and shed from the plasma membrane [60]. Interestingly, Malleret et al., observed a very rapid loss of TfR upon P. vivax invasion, suggesting a substantial increase in reticulocyte maturation rate, although the mechanism (e.g. enhanced exosome release) for TfR loss upon invasion remains unclear. It also remains to be seen if such rapid remodeling occurs ex vivo and if it is a general feature of Plasmodium spp. infection or unique to P. vivax.

Heterogeneity in P. vivax tissue distribution.

There has been an increasing recognition of the importance of deep tissue reservoirs for malaria parasites, and recent work has highlighted the role of infections in the bone marrow. Recent molecular and autopsy studies have identified P. falciparum in the extravascular space of the bone marrow, where there was a significant enrichment for early gametocyte stages which are subsequently thought to transmigrate into the peripheral circulation [65,66]. Furthermore a recent study with the rodent malaria parasite P. berghei has demonstrated both the existence of a cryptic asexual cycle within the bone marrow and commitment to gametocytes formation upon invasion of early reticulocytes [67].

P. vivax has been observed in human bone marrow [68], suggesting that this tissue may host active P. vivax infections. There is also increasing evidence that the bone marrow is enriched in P. vivax gametocytes [69]. Recently, an autopsy study performed on P. vivax-infected Aotus and Saimiri non-human primates demonstrated that a large proportion of asexual- and sexual-stage parasites (~ 30%) were found in the bone marrow parenchyma, strongly supporting the role of this organ as a niche for P. vivax [70].

Many unresolved questions remain regarding the tissue tropism of P. vivax including whether parasite asexual replication takes place in the bone marrow extravascular space, or whether parasites accumulate via transmigration. Do parasites distinguish between bone-marrow reticulocytes and peripheral reticulocytes and if so, what are the molecular determinants and is it linked to gametocyte commitment? The invasion of P. vivax appears to demonstrate similarities with, but significant differences from that of other Plasmodium spp. Reticulocyte tropism could open the window to our understanding Plasmodium infection of deep tissue niches.

Highlights.

• P. vivax has multiple and novel invasion ligand families (PvDBPs, PvRBPs, PvTRAgs)

• PvDBP binds to DARC

• PvRBP2b binds to the reticulocyte-tropic transferrin receptor (TfR/CD71)

• P. vivax can infect reticulocytes in tissue niches including bone marrow

Terms:

DARC

Duffy Antigen Receptor for Chemokines

EBP

erythrocyte binding proteins

RBP

reticulocyte binding proteins

TRAg

tryptophan rich antigens

TfR/CD71

transferrin receptor

Invasion ligand/adhesin

parasite-expressed protein that binds to host surface membrane protein mediating host tropism during invasion

Host receptor

host surface membrane protein targeted by parasite invasion ligands