Babesia parasites modify the host red blood cell

Babesia species are protozoan parasites belonging to the phylum Apicomplexa and are transmitted by ticks. More than 100 Babesia species have been reported, which infect mammalian and avian hosts, and several species are known to parasitize domestic animals such as cattle, horses, sheep, goats, as well as dogs [1]. Bovine Babesia species (B. bovis, B. bigemina, B. divergens, B. major, and B. ovata) capable of mediating disease are widely distributed in temperate and tropical regions in the world. It is estimated that 1.2 billion cattle are at risk of infection, and bovine babesiosis is a major cause of economic loss within the beef and dairy industries, thus highlighting the significance of this protozoan parasite in veterinary medicine [2]. Several Babesia species, including Babesia microti and B. divergens, have gained attention as pathogenic species for emerging zoonotic diseases [3].

Babesia sporozoite stage parasites are released from the salivary glands of infected vector ticks during a blood meal and directly invade red blood cells (RBCs) to become pear-shaped “pyriform” intraerythrocytic stage parasites [4,5,6]. The pyriform parasites rapidly replicate by binary fission, leading to paired parasites inside the RBC. Some clades of Babesia, such as B. microti and B. duncani, form 4 merozoites per division, which result in the appearance of the Maltese cross form [7]. This multiplication and the following destruction of the host RBC during parasite egress lead to pathologies including fever, anemia, jaundice, and hemoglobinuria. Although most apicomplexan parasites infect nucleated host cells, Babesia and a few closely related genera such as the malaria parasite Plasmodium have evolved to parasitize enucleated RBCs. During invasion and subsequent development, the parasite modifies host RBCs by exporting proteins. Modification of nucleated host cells is known for many apicomplexan parasites, and exported proteins manipulate host cell processes such as transcription and kinase activity [8]. Babesia and Plasmodium reside inside anucleate RBCs (Plasmodium also has a liver stage, but Babesia solely multiply inside RBCs), and, therefore, the role of exported proteins is likely different from apicomplexans, which infect nucleated cells. The repertoire of exported proteins is called the “exportome”, and in the instance of Plasmodium, hundreds of such proteins have been identified within numerous expanded gene families [9,10]. Extensive studies have described the role of Plasmodium exportome proteins in reinforcing the RBC cytoskeleton, altering the RBC surface to evade immune recognition, inducing cell adhesive properties, and changing the RBC membrane permeability to allow the uptake of nutrients and release of parasite metabolic waste products [11]. Since the exportome is important for the parasite survival and its pathogenicity, these proteins are considered to be drug or vaccine targets [11].

Despite the similarities to Plasmodium infection, less is known regarding the Babesia exportome, and the degree of overlap, if any, in the repertoire of exported proteins. While it is known that infected RBCs (iRBCs) are modified by Babesia parasites, such as to induce RBC adhesive properties and alter permeability, prior to the past few years, only a few exported proteins were identified and functional analysis was scarcely done [12,13]. However, the recent completion of genome sequences and the establishment of transfection systems for several Babesia species has aided characterization of the exportome, and the functions of several exported proteins have been partially revealed.

Babesia exportome and the motifs responsible for protein export

Before the identification of the exportome in Babesia parasites, modification of iRBC by the parasite was known from studies on cerebral babesiosis, a fatal complication in cattle infected with the most virulent bovine Babesia species, B. bovis [14]. Sequestered iRBCs were found in the cerebral microcapillary of infected cattle, mediated by the cytoadhesion of iRBCs to brain microcapillary endothelial cells. Plasmodium falciparum, a causative agent of human malaria, produces similar clinical symptoms, called cerebral malaria [15]. B. bovis and P. falciparum iRBCs both show protrusions on the surface of the RBC, which, based upon their topology, are called ridges in B. bovis iRBCs (Fig 1A) and knobs in P. falciparum iRBCs [16,17]. During and following invasion, Plasmodium proteins are secreted from rhoptries and dense granules into the parasitophorous vacuole and are trafficked to the iRBC cytoplasm and surface [11]. Several proteins responsible for knob formation have been identified [11]. Many exported proteins are regulated by a host targeting signal “PEXEL-motif”, which consists 5 amino acids RxLx(x)E/Q/D (x represents any amino acids) located downstream of a signal peptide that mediates translocation to the endoplasmic reticulum (ER) [18,19]. Within the ER, PEXEL-containing proteins are proteolytically cleaved at the PEXEL motif and the new N-terminus is recognized for targeting to the parasitophorous vacuole and translocation across the parasitophorous vacuole membrane (PVM) by a Plasmodium-specific translocon [20,21]. Some of the proteins in the RBC compartment are transported to vesicle like structures of parasite origin within the RBC cytoplasm, termed Maurer’s clefts in P. falciparum, for either final residence or subsequent delivery to a functional destination within the RBC [10]. In addition, P. falciparum exports several PEXEL-negative proteins (PNEPs) [22]. PNEPs typically contain a transmembrane domain or N-terminal signal sequence, and these proteins are trafficked either in a soluble or membrane-bound form [23,24].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Electron microscopy images of B. bovis-infected RBCs.

Scanning electron microscopy (A) and TEM (B) and (C) of B. bovis-iRBC binding to bovine brain endothelial cells. Membranous structures are seen in the TEM image and indicated by a white arrow. The red arrows show the ridges mediating binding of iRBCs to endothelial cells. Scale bar = 1 μm. iRBC, infected RBC; RBC, red blood cell; TEM, transmission electron microscopy.

https://doi.org/10.1371/journal.ppat.1010770.g001

Unlike Plasmodium parasites, Babesia and its close relative Theileria break down their PVM within minutes after completion of invasion [25–27]. The lack of a PVM precludes the need for a translocon for the export of proteins across this membrane. Maurer’s cleft-like structure have not been identified in Babesia parasites; however, membranous structures have been observed in electron microscopy images and are proposed to be either vesicles responsible for protein trafficking to the RBC or the remains of the dissociated PVM ([25,27]; Fig 1B). Due to the structural differences of RBCs modified by Babesia versus Plasmodium, it is plausible that the export machinery is simpler for Babesia and that motifs mediated trafficking differ between the 2 parasite genera.

The first identified Babesia exportome protein was the B. bovis variant erythrocyte surface antigen 1 (VESA1) ([28]; Fig 2). It was known that antibodies raised against B. bovis iRBC can mediate their agglutination, leading to speculation of parasite-encoded antigens on the RBC surface [29]. A family of highly antigenic variable proteins were later identified on the iRBC surface and named VESA1 [30]. Currently, it is known that VESA1 is the product of a multigene family and localizes on the ridges of the iRBCs, discussed further below [31].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Immunofluorescence microscopy of transgenic parasites expressing myc-tagged exported protein.

Myc-tagged proteins were episomally expressed in B. bovis (VESA1, SBP3, SmORF, BbVEAP, Bbmtm, and BbTPR-related) or B. ovata (BoMFS). The expression plasmid was constructed and transfected by electroporation as described [32,33], and transgenic parasites were selected using WR99210. Thin smears of cultured parasites were prepared for indirect immunofluorescence microscopy, fixed with a 1:1 acetone:methanol mixture, and reacted with anti-myc mouse monoclonal antibody (9B11, Cell Signaling) followed by Alexa Fluor 488-conjugated goat anti-mouse IgG (Invitrogen, green). Nuclei were stained with Hoechst 33342 (Hoechst, blue). Scale bar = 5 μm. BbVEAP, B. bovis VESA1-export associated protein; IgG, immunoglobulin G; SBP3, spherical body protein 3; SmORF, Small Open Reading Frame; VESA1, variant erythrocyte surface antigen 1.

https://doi.org/10.1371/journal.ppat.1010770.g002

Another Babesia-exported protein identified from early studies was an immunodominant B. bovis protein, spherical body protein 1 (SBP1, originally named Bb-1) [34]. Immunofluorescence assays revealed that SBP1 was detected at the permeabilized iRBC membrane and inside the parasite but not using preparations of unfixed iRBC, indicating that SBP1 was localized at the cytoplasmic side of iRBCs [35]. Immunoelectron microscopy analysis revealed SBP1 within the spherical bodies of the parasites, which are unique organelles for Babesia and Theileria and appear to be analogous to dense granules of other apicomplexan parasites [36]. Antibodies raised against spherical bodies identified 2 more proteins, SBP2 and SBP3 ([36,37]; Fig 2). Similar to SBP1, SBP2 and SBP3 localized on the cytoplasmic face of iRBCs and spherical bodies of the parasites. Currently, a fourth protein (SBP4) has been deposited in GenBank, and the localization in iRBCs has been confirmed [38,39]. Although SBPs contain a signal peptide and PEXEL-like motif (PLM), they are heterogeneous in structure; thus, future functional characterizations might inspire a less generic renaming of these proteins. While the export of VESA1 and SBPs were shown experimentally, a PEXEL motif was not found in these proteins.

To date, 13 proteins including 5 protein families have been experimentally confirmed to be exported by Babesia (Table 1 and Fig 3). The completion of the genome sequence of B. bovis enabled bioinformatic approaches to find new exported proteins [40]. For example, Gohil and colleagues selected exportome candidates by virtue of their possessing a signal peptide, transmembrane domain, GPI anchor, and domains/features of extracellular proteins [41]. They found 214 proteins with signal peptides but without transmembrane domains or GPI anchors and concluded with the identification of 3 novel candidate exportome proteins (Gene IDs: BBOV_II007340, BBOV_II004730, and BBOV_IV009240). Another study used a bioinformatic approach to describe a PLM that is consistent with an exportome prediction [42]. The PEXEL-related signal was originally identified from another apicomplexan parasite, Toxoplasma gondii and named PLM [43]. This motif consists of Rx(x)L and is cleaved by an aspartyl protease, similar to Plasmodium parasites. PLM is required for Toxoplasma protein trafficking to dense granules and PVM binding. Pelle and colleagues searched for PLM in B. bovis proteins and found that a Rx(x)L motif can be found at a higher frequency than permutated control sequences. PLM preceded by a signal peptide can be found in SBPs and Small Open Reading Frame proteins (SmORF) (Fig 2). Using the SmORF or SBP2t11 (a protein coded by one of the truncated copies of SBP2) N- terminal sequences fused to a fluorescence protein marker, they demonstrated that a signal sequence is sufficient to export the fluorescence reporter protein to the iRBC. A mutated PLM resulted in a reduced retention of proteins within the spherical bodies, suggesting that the PLM works as a retention signal in spherical bodies for proper protein maturation and release and thus serves a different function than the PEXEL system in Plasmodium [42]. Additionally, these proteins were proteolytically processed at the signal sequence and PLM [42]. When SmORF with a complete PLM was expressed in Plasmodium, the protein was proteolytically processed and exported across the PVM to the RBC [42]. These results suggest that even though there is a functional difference in the PLM/PEXEL motif between these parasites, the trafficking pathway of Babesia PLM-positive proteins is related to PEXEL-dependent trafficking in Plasmodium [42].

Another recent bioinformatic approach used prediction by machine learning [44]. This study used Gohil’s 214 proteins as an export-positive training dataset and predicted an exportome for several Babesia species. Using 3 prediction algorithms, they estimated an additional 144 potential exportome proteins for B. bovis; and 371 and 196 exportome proteins for B. bigemina and B. canis, respectively (with probability more than 70%). This study was done in silico, and, therefore, confirmatory cellular studies are necessary. In summary, advances in computational biology have given new opportunity to describe the Babesia exportome.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Experimentally confirmed Babesia exportome proteins.

https://doi.org/10.1371/journal.ppat.1010770.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Babesia iRBC.

Babesia parasites export numerous proteins to remodel the host RBC using secretory pathway. Several of these proteins are routed through spherical bodies, while others are either directly exported to RBC or deposited in unknown secretory organelles. B. bovis produces ridges on the surface of iRBCs to express VESA1 that mediates sequestration and immune evasion. Of spherical body proteins, SBP1, SBP2, SBP3, and SBP4 are associated with the RBC membrane; VEAP is released to the RBC cytoplasm and is essential for parasite growth and VESA1 export; mtm integrates into the RBC membrane and is likely responsible for nutrient uptake. Vesicular structures in the RBC are produced following parasite invasion that might have a role in protein export. Regarding other Babesia spp., VESA1 expression in iRBC was shown for B. orientalis and MFS export to RBC for B. ovata, though their functions remain to be determined. iRBC, infected red blood cell; mtm, multi-transmembrane protein; PPM, parasite plasma membrane; RPM, RBC plasma membrane; SBP1, spherical body protein 1; SmORF, Small Open Reading Frame protein; VEAP, VESA1-export associated protein; VESA1, variant erythrocyte surface antigen 1.

https://doi.org/10.1371/journal.ppat.1010770.g003

Immune evasion and cytoadhesion

To establish an infection in the host, Babesia parasites must evade immune recognition, including avoidance of an antibody-mediated immune response or direct clearance of iRBCs in the spleen. The spleen plays a critical role in recognizing and clearing abnormal and aged RBCs, as well as RBCs altered due to infection. In the case of B. bovis, avoidance of splenic clearance is achieved by the display of parasite-encoded ligands on the surface of the iRBCs, which confer capillary endothelial adhesion and sequestration in internal organs [45]. However, in this manner, the parasite exposes itself to a possible antibody-mediated immune response, which it in turn evades over the course of an infection by sequentially switching the expression of the surface ligands with other members of a repertoire of antigenically variant surface proteins. Antigenic variation in Babesia parasites was first described in the rodent parasite B. rodhaini [46], and following in B. bigemina [47], and was suggested to occur in B. bovis by the parasite-encoded proteins on the surface of iRBC that were named VESA1 [31,48,49]. VESA1 is a heterodimeric protein encoded by the multicopy ves1α and ves1β gene families in B. bovis [31,50]. The first genome sequence of B. bovis revealed 119 copies of ves1 genes [40], which, with improvement of genome assembly, increased to 133 genes consisting of 81 ves1α, 48 ves1β, and 4 unclassified ves1 [51]. In the B. bovis genome, ves1α and ves1β gene pairs are located in a divergent orientation and are expressed at a site called locus of active transcription (LAT) [52]. It is believed that ves1 genes show monoallelic expression [53] in which an active LAT containing a ves1α and ves1β pair are simultaneously expressed by a bidirectional promoter located within the LAT intergenic region [54]. Antigenic variation and the resulting immune evasion in B. bovis occur through epigenetic in situ switching of transcribed ves1 and genomic recombination, which results in mosaicism of ves1 genes [55].

Sequestration of B. bovis is mediated by interaction of iRBC surface ridges with bovine endothelial cells (Fig 1C) [16,56,57]. The sequestered iRBCs can disrupt blood flow in internal organs and in the brain, which causes cerebral babesiosis [16]. It was proposed that VESA1 is responsible for cytoadhesion of iRBCs based upon the observations that VESA1 proteins are clustered on ridges, and specific monoclonal antibodies against VESA1 inhibited binding of iRBCs to endothelial cells and reversed binding of cytoadhered iRBCs [57]. Additionally, chemical disruption of VESA1 export or trypsin treatment to remove surface proteins resulted in inhibition of binding, which was regained when VESA1 was repopulated on the RBC [56]. Further transcriptomics analysis supported a VESA1 role in pathogenesis [58]. VESA1 has a single transmembrane domain, a large extracellular domain, and a short cytoplasmic tail [31,50], a structure that superficially resembles the P. falciparum RBC surface adhesin, erythrocyte membrane protein 1 (PfEMP1). The extracellular region of VESA1 proteins possess a cysteine- and lysine-rich domain and variant domain conserved sequences [31,59]. While the role of PfEMP1, the specific binding domains, and the host receptors are well documented for cerebral and pregnancy-associated malaria [60–62], the role of binding domains, receptors, and the impact of VESA1 for sequestration remain to be determined for Babesia.

Two additional exported proteins were shown to be involved in cytoadhesion of B. bovis iRBC, SBP2t11, and VESA1 export-associated protein, BbVEAP ([33,63]; Fig 2). Up-regulation of SBP2t11 was associated with low virulence of B. bovis, and its overexpression reduced binding of iRBCs to endothelial cells [63,64]. While the cleavage at PLM and the export of SBP2t11 has been demonstrated [63], it is likely involved in cytoadhesion indirectly, as it is not expressed on the surface of iRBC. It remains unclear whether SBP2t11 affects VESA1 export, ridge formation, or other factors responsible for cytoadhesion. Knockdown of BbVEAP using the glmS ribozyme system disrupted the export of VESA1, decreased ridge numbers, and abrogated cytoadhesion of iRBCs [33]. BbVEAP may function as a chaperone for the export of VESA1 as an integral protein and ridge-forming proteins; however, immunoprecipitation of BbVEAP did not confirm a direct interaction with VESA1 [33]. Given that BbVEAP knockdown did not affect SBP4 export, BbVEAP expression is necessary for the export and correct localization of a subset of proteins including VESA1 and ridge-forming proteins. Additionally, it was shown that BbVEAP is indispensable for parasite development in the RBC, making it the only known essential exportome protein [33,65]. The existence of VEAP among piroplasma parasites indicates a piroplasma-specific conserved function. The essentiality of VEAP could be due to its role in the export of other essential proteins such as channels or transporters, a hypothesis that needs future confirmation.

Although ves1α and ves1β genes are unique for B. bovis, ves-like genes are found in homologous genomic regions in all Babesia species for which genome sequence is available (Fig 4A). Ves1a and ves1b genes are found in B. bigemina, B. divergens, and B. ovata and encode proteins that contain a single transmembrane domain and a short cytoplasmic region at C-terminus similar to the products of ves1α and ves1β genes of B. bovis [66,67]. A shorter ves gene group called ves2 encode proteins lacking the C-terminal transmembrane domain and the cytoplasmic region and were found based on their homology to the 5′ end of ves1 in B. bigemina, B. divergens, and B. ovata [66,67]. The expanded ves2 genes of B. ovata cluster together with B. bigemina ves2 based upon high sequence identity (Fig 4A). Given the fact that ves2 exists in the homologous positions of smorf, it was suggested that ves2 is analogous to smorf in these parasites [66]; however, the function of both gene families is unknown. VESA1 from B. bovis was experimentally confirmed to be expressed on the iRBCs surface [28], and VESA1 from B. orientalis was shown to be exported to iRBC [68]. While in B. bovis VESA1 is responsible for cytoadhesion and antigenic variation [69], cytoadhesion has not been documented for other Babesia spp., despite the presence of expanded ves gene families in their genomes. It is likely that VESA from other Babesia species are also surface proteins responsible for antigenic variation [66]. Future experiments are needed to characterize the functions of VESA in Babesia spp. in the mammalian host and tick vector, as the expression of some ves1 genes were up-regulated in the kinetes, the invasive stage of parasites in the tick hemolymph, of B. bovis [51].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Homology clustering based on sequence similarities of ves and multi-transmembrane protein encoding genes.

(A) ves and smorf genes sequence were extracted from piroplasmaDB (https://piroplasmadb.org/piro/app), and a sequence similarity network was visualized by Gephi. (B) The figure was reproduced from Hakimi and colleagues [33]. The genes encoding proteins with more than 8 TM domains were extracted from piroplasmaDB and clustered. SmORF, Small Open Reading Frame; TM, transmembrane.

https://doi.org/10.1371/journal.ppat.1010770.g004

Nutrient uptake

Mammalian RBCs are terminally differentiated and anucleate and lack cellular machinery and metabolism. Babesia parasites have a fast metabolism and require nucleic acids and proteins, and the relatively short doubling time (10 to 12 h for B. bovis) and DNA replication (2.9 h for B. bovis) [65] necessitate uptake of metabolic precursors from the host plasma, such as sugars, purines, and amino acids. For P. falciparum iRBCs, studies using osmotic lysis, patch-clamp methods, and radioactive labeled substrates have documented increases in permeabilities to various nutrients, called the new permeability pathway or plasmodial surface anion channel (PSAC) [70,71]. Similarly, it has been reported that Babesia parasites increase RBC permeability to several solutes to facilitate the uptake of essential nutrients for their growth. Sorbitol uptake by RBCs infected with B. divergens and B. bovis was shown through osmotic lysis and transmittance assay and was used for the enrichment of iRBCs [33,72]. The uptake of glucose, nucleosides, and several amino acids was shown for several Babesia spp. [72–77]. While PSAC is an anion channel, the temperature dependence and relatively slower uptake of solutes by Babesia in comparison to malaria parasite suggest a carrier-type nature of the putative channels [72,78].

PSAC activity is determined by the CLAG3 protein [79,80], which is released during merozoite invasion. Orthologs of this protein appear to be absent in Babesia parasites. Recently, a protein encoded by a multicopy gene family (Fig 4B) with a multi-transmembrane structure (mtm) was identified in B. bovis using proteomics of the iRBC membrane ([33]; Fig 2). The expression level of mtm correlated with blasticidin S (BS) resistance, suggesting a role in BS uptake [33]. While several mtms are expressed simultaneously, only the expression of a single mtm was linked with BS resistance [33]. This may indicate that each mtm is involved in the transport of a specific substrate or they have multiple functions. BS resistance in P. falciparum results in a strong impairment of PSAC activity, as evidenced by a significant reduction of sorbitol uptake and being refractory to synchronization using sorbitol [81], while in B. bovis, sorbitol uptake and RBC lysis were slightly delayed. This suggests that while PSAC activity is determined by a single CLAG protein, the increase in RBC permeability by B. bovis is determined by a combination of several mtms. The observation of the simultaneous expression of a subset of mtms supports this hypothesis.

Although mtm genes only exist in B. bovis and Babesia sp. Xinjiang, expanded gene families that encode proteins with multi-TM structures are found in all piroplasms (Fig 4B; [33]). B. bigemina, B. ovata, and B. divergens have an expanded major facilitator superfamily mfs, while tpr (Theilaria parva repeat), which was initially discovered in T. parva, is expanded in Theileria spp. (Figs 2 and 4B; [82,83]). Whether mtm or other expanded multi-TM gene families are responsible for host cell permeability by piroplasms remains to be determined. Babesia cabali, the causative agent of equine piroplasmosis, produces tubular structures in the cytoplasm of infected RBCs, which appear to connect the parasite to the host serum [84]. It is not known if these tubular structures are responsible for nutrient uptake as a direct “duct”, and the genes responsible for the generation of these microtubules have not been determined.