*Article Babesia microti* **Immunoreactive Rhoptry-Associated Protein-1 Paralogs Are Ancestral Members of the Piroplasmid-Confined RAP-1 Family**

**Reginaldo G. Bastos 1,\* , Jose Thekkiniath <sup>2</sup> , Choukri Ben Mamoun <sup>3</sup> , Lee Fuller <sup>2</sup> , Robert E. Molestina <sup>4</sup> , Monica Florin-Christensen 5,6 , Leonhard Schnittger 5,6 , Heba F. Alzan 1,7,8 and Carlos E. Suarez 1,9,\***


**Abstract:** *Babesia*, *Cytauxzoon* and *Theileria* are tick-borne apicomplexan parasites of the order Piroplasmida, responsible for diseases in humans and animals. Members of the piroplasmid rhoptryassociated protein-1 (pRAP-1) family have a signature cysteine-rich domain and are important for parasite development. We propose that the closely linked *B. microti* genes annotated as BMR1\_03g00947 and BMR1\_03g00960 encode two paralogue pRAP-1-like proteins named BmIPA48 and Bm960. The two genes are tandemly arranged head to tail, highly expressed in blood stage parasites, syntenic to *rap-1* genes of other piroplasmids, and share large portions of an almost identical ~225 bp sequence located in their 50 putative regulatory regions. BmIPA48 and Bm960 proteins contain a N-terminal signal peptide, share very low sequence identity (<13%) with pRAP-1 from other species, and harbor one or more transmembrane domains. Diversification of the piroplasmid-confined *prap-1* family is characterized by amplification of genes, protein domains, and a high sequence polymorphism. This suggests a functional involvement of pRAP-1 at the parasite-host interface, possibly in parasite adhesion, attachment, and/or evasion of the host immune defenses. Both BmIPA48 and Bm960 are recognized by antibodies in sera from humans infected with *B. microti* and might be promising candidates for developing novel serodiagnosis and vaccines.

**Keywords:** *Babesia microti*; BmIPA48; BMR1\_03g00960; piroplasmid rhoptry-associated protein-1 (pRAP-1); human babesiosis

#### **1. Introduction**

*Babesia*, *Cytauxzoon* and *Theileria* are tick-borne apicomplexan piroplasmid parasites of vertebrates that invade and reproduce asexually in erythrocytes. These parasites are a major concern to human and animal health and cause an important economic burden worldwide. *Babesia* parasites are responsible for acute and persistent hemolytic disease in several wild and domestic vertebrate species, including human. While *Theileria* parasites are transmitted transstadially by ticks, *sensu stricto* (s.s.) *Babesia* spp. are transovarially and,

**Citation:** Bastos, R.G.; Thekkiniath, J.; Ben Mamoun, C.; Fuller, L.; Molestina, R.E.; Florin-Christensen, M.; Schnittger, L.; Alzan, H.F.; Suarez, C.E. *Babesia microti* Immunoreactive Rhoptry-Associated Protein-1 Paralogs Are Ancestral Members of the Piroplasmid-Confined RAP-1 Family. *Pathogens* **2021**, *10*, 1384. https://doi.org/10.3390/ pathogens10111384

Academic Editors: Estrella Montero, Jeremy Gray, Cheryl Ann Lobo and Luis Miguel González

Received: 17 September 2021 Accepted: 23 October 2021 Published: 26 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

in some species, also transstadially, transmitted. Other piroplasmids, such as *B. microti*, are defined as *sensu lato* (s.l.) *Babesia* parasites, based on their transstadial mode of transmission and the absence of schizont stages in their life cycles [1–3].

Human babesiosis is an emergent worldwide zoonosis caused by several *Babesia* spp., including the s.s. *B. divergens* and the s.l. *B. microti*, the latter of which is the predominant agent in the Northeastern and Midwest regions of the US [4,5]. As for other piroplasmids, the life cycle of *B. microti* is dixenic, involving an invertebrate definitive host and a vertebrate host. In the US, the primary vertebrate host is the white-footed mouse (*Peromyscus leucopus*) and the invertebrate host is a tick of the genus *Ixodes*, such as *I. scapularis*. However, humans are accidental and dead-end hosts when bitten by infected ticks. Importantly, human-to-human transmission of *B. microti* may occur via contaminated blood transfusions [6,7]. Due to climate change and human activity, the geographic distribution of *I. scapularis*, and hence of *B. microti*, is expanding rapidly in the US [8]. In addition, the finding of vertical transmission in the white-footed mouse indicates a potentially relevant way of parasite dissemination without the participation of the tick vector [9]. The disease caused by *B. microti* in humans may vary from asymptomatic or subclinical to acute and chronic manifestations, which can be lethal in immunocompromised patients. Clinical manifestations of acute human babesiosis include fever, hemolytic anemia, acute respiratory distress and multiorgan dysfunction [10]. Because of the expansion of the tick habitat and the constant increase in cases of human babesiosis in the US, there is a need to develop vaccines and improved diagnostics against *B. microti*, which requires identification of conserved immunogenic proteins in this apicomplexan parasite.

Apicomplexan parasites, including *B. microti*, are equipped with an apical complex with at least three distinct secretory organelles known as the rhoptries, micronemes, and spherical bodies or dense granules. These organelles play an essential role in host cell invasion by the parasite [11]. Once the parasite is committed to invasion, it is quickly and actively propelled inside the target cell by the activity of an actin motor, with intervention of the cytoskeletal structures of the parasite [12]. Rhoptry proteins are probably involved in the formation of the parasitophorous vacuole (PV), a membranous structure separating the parasite from the cytoplasm of the host cell, that disappears quickly upon invasion in *Babesia* parasites [13]. Remarkably, *B. microti* also developed a mechanism for vesiclemediated antigen export generating an interlacement of vesicles which extends from the plasma membrane of the parasite into the cytoplasm of the host erythrocyte [14]. Few rhoptry proteins have been so far identified and characterized in *Babesia* parasites. Initial studies performed mainly in *B. bovis* and *B. bigemina* were focused on the functional role of rhoptry-associated protein-1s (RAP-1s), which were later identified in all piroplasmids, including other *Babesia* spp., *Theileria* spp. and *Cytauxzoon felis* [15–23]. We hereby refer to these proteins as piroplasmid RAP-1s (pRAP-1s). It is possible that the function of these piroplasmid-specific proteins is needed to support unique features of the parasite life cycle, such as parasite-attachment to the erythrocyte, dissolution of the PV in *Babesia*, the zippermediated invasion of *Theileria*, or other events that may be related to erythrocyte invasion and egress [24,25]. The *prap-1* gene superfamily encodes the paralogs *rap-1* and RAP-1 related antigens (*rra*) in *B. bovis* [26]. Plasmodial RAP-1 shares the same denomination with pRAP-1s, but they are unrelated non-homologous proteins [27]. Since pRAP-1 proteins are highly immunogenic and can be targeted for neutralization-sensitive antibodies, they may be attractive candidates for diagnostic assays or subunit vaccines against *Babesia* and *Theileria* parasites [16,28–34].

The piroplasmid-specific RAP-1 family domain (PF03085) contains a characteristic motif of four cysteine (Cys) residues and a single conserved tyrosine (Tyr) residue. Other definitions of the members of this protein family are based on localization or function, which are still waiting experimental confirmation. Although the pRAP-1 proteins have been identified and annotated in genomes of *Babesia* spp. s.s., *Cytauxzoon felis*, and *Theileria* spp. parasites, they remain not fully identified in the genome of the s.l. parasite *B. microti*. A RAP putative protein (XP\_021337499) was annotated in the genome of *B. microti*, but this

protein, which is homologous to *Plasmodium* and *Toxoplasma* RAPs, lacks the characteristic motifs of the members of the *Babesia*/*Theileria* pRAP-1 superfamily [35]. Thus, the presence of canonical *Babesia*/*Theileria* pRAP-1 genes has yet to be reported in *B. microti*. We hypothesized that, like *Babesia* and *Theileria* parasites, the genome of *B. microti* also includes genes encoding for pRAP-1 or pRAP-1-like proteins. Furthermore, because of the relatively distant phylogenetic relationship of *B. microti* with piroplasmid parasites such as *Babesia* s.s. and *Theileria* s.s. [1], we propose that the pRAP-1-like proteins encoded by *B. microti* may have diverged dramatically from the pRAP-1 molecules expressed in other piroplasmids, resulting in a low non-significant sequence identity, but conservation of important structural features. Indeed, neither a common BLASTp nor a Pfam search resulted in hit or domain report, respectively. Therefore, we carried out alternative search strategies on the *B. microti* genome based on the previously detected conserved synteny in the genome regions of piroplasmid parasites where the *prap-1* loci are encoded and found two head-to-tail oriented linked genes, BMR1\_03g00947 and BMR1\_03g00960, encoding for proteins with structural characteristics that are compatible with the pRAP-1 molecules. Although the database searches did not result in hits, synteny analysis and the presence of highly conserved amino acid residues of structural importance organized as the Cys-rich domains of the pRAP-1s proteins strongly suggest that the presented two genes encode for pRAP-1 homologs in *B. microti*. Since these putative *B. microti* pRAP-1 proteins lack significant sequence identity with pRAP-1 domains of other pRAP-1s, we designated them pRAP-1-like proteins. For the aforementioned reasons, *B. microti* pRAP-1-like proteins have previously remained unnoticed, though these proteins have been identified and shown to be expressed in *B. microti* merozoites [35,36].

#### **2. Results**

#### *2.1. Two Tandemly Arranged RAP-1 Syntenic Genes of B. microti Encode Proteins Containing Non-Canonical Piroplasmid RAP-1 Cys-Rich Domains*

The piroplasmid RAP-1 proteins contain a characteristic Cys-rich domain, signal peptide, and other short conserved sequence motifs. The salient features of some typical pRAP-1 and RRA representatives of this family are schematized in Figure S1. In this study, we first searched the predicted proteome of the *B. microti* R1 strain for the identification of proteins containing pRAP-1 Cys motifs using Delta-Blast analysis against a query of the *B. bigemina* RAP-1c Cys-rich domain (CLGSKDEHHCASQIAAYVARCKE), also typical for the pRAP-1s of *B. bovis* (Figure 1). This search revealed a hit with the hypothetical protein encoded by gene BMR1\_03g00960, here referred to as Bm960 (Figure 1). This finding prompted us to investigate the corresponding gene locus for the presence of other *rap-1* related genes and for synteny with *B. bovis*, *B. bigemina* and *T. equi rap-1* loci. Sequence analysis revealed that the BMR1\_03g00947 gene, reffered to as BmIPA48 (Figure S1), located immediately next to Bm960, and separated by an 800-bp intergenic region, encodes for a protein also containing a similar RAP-1-like Cys-rich region, including a key conserved Tyr residue in its amino terminal (Figure 1, Figure 2B,C and Figure S2).

*Pathogens* **2021**, *10*, x FOR PEER REVIEW 4 of 16

face of the parasite, as previously predicted [35].

face of the parasite, as previously predicted [35].

typical residues of the pRAP-1 proteins in the NT-region of the molecules, as well as other short amino acid motifs (Figure S2). The protein encoded by gene BMR1\_ BmIPA48 also contains a series of tandem repeats in its C-terminal region, a feature that is shared with the *B. bovis* RAP-1 proteins (Figures 3 and S3). Strikingly, sequence analysis of the noncoding regions immediately upstream of genes BmIPA48 and Bm960 revealed conservation of a 300-bp sequence (Figures 3 and S4), suggesting that expression of these two proteins might be coordinated, despite their non-relatedness in sequence. In addtion, secondary structure sequence analysis performed in silico using TMpred suggests that BmIPA48 contains a signal peptide (aa 4–24) and a putative transmembrane (TM) region (aa 164–186) (Figure 4). Since no TM domains were previously reported in this protein, the prediction was confirmed using the alternative algorithm Phobius, which also showed the presence of a TM domain in the same region (Figure S6). Bm960 protein also contains a predicted signal peptide, two TM domains, and lack a predicted GPI anchor attachment site (Figure 4). Collectively, these features are fully consistent with expression on the sur-

typical residues of the pRAP-1 proteins in the NT-region of the molecules, as well as other short amino acid motifs (Figure S2). The protein encoded by gene BMR1\_ BmIPA48 also contains a series of tandem repeats in its C-terminal region, a feature that is shared with the *B. bovis* RAP-1 proteins (Figures 3 and S3). Strikingly, sequence analysis of the noncoding regions immediately upstream of genes BmIPA48 and Bm960 revealed conservation of a 300-bp sequence (Figures 3 and S4), suggesting that expression of these two proteins might be coordinated, despite their non-relatedness in sequence. In addtion, secondary structure sequence analysis performed in silico using TMpred suggests that BmIPA48 contains a signal peptide (aa 4–24) and a putative transmembrane (TM) region (aa 164–186) (Figure 4). Since no TM domains were previously reported in this protein, the prediction was confirmed using the alternative algorithm Phobius, which also showed the presence of a TM domain in the same region (Figure S6). Bm960 protein also contains a predicted signal peptide, two TM domains, and lack a predicted GPI anchor attachment site (Figure 4). Collectively, these features are fully consistent with expression on the sur-

**Figure 1.** Representative figure of the comparisons performed with the pRAP-1 Cys-rich motif among babesial rRAP-1 proteins and the newly identified *B. microti* putative pRAP-1. The comparisons include the Cys-rich regions of *B. bovis* RRA, *B. bovis* RAP-1, *B. bigemina* RAP-1c, and the *B. microti* RAP-1- like proteins BmIPA48 and Bm960. **Figure 1.** Representative figure of the comparisons performed with the pRAP-1 Cys-rich motif among babesial rRAP-1 proteins and the newly identified *B. microti* putative pRAP-1. The comparisons include the Cys-rich regions of *B. bovis* RRA, *B. bovis* RAP-1, *B. bigemina* RAP-1c, and the *B. microti* RAP-1- like proteins BmIPA48 and Bm960. **Figure 1.** Representative figure of the comparisons performed with the pRAP-1 Cys-rich motif among babesial rRAP-1 proteins and the newly identified *B. microti* putative pRAP-1. The comparisons include the Cys-rich regions of *B. bovis* RRA, *B. bovis* RAP-1, *B. bigemina* RAP-1c, and the *B. microti* RAP-1- like proteins BmIPA48 and Bm960.

**Figure 2.** Synteny map of the *rap-1* locus of *T. equi*, putative *rap-1 B. microti*, and typical s.s. *B. bovis*. (**A**) Structure of the *prap-1* and *rra* genes localized in the chromosome 4 of *B. bovis*. (**B**) Conserved synteny among the *rap-1* loci of *B. bovis* and *T. equi* and the BMR1\_03g00960 (BmIPA48) and BMR1\_03g00947 (Bm960) genes of *B. microti*, that encode for proteins containing the typical Cys-rich region of the pRAP-1 proteins. (**C**) Conserved synteny among the *B. bovis rra* and the *B. microti* BMR1\_03g00960 and BMR1\_03g00947 genes.

> As shown in Figure 2A, the chromosome 4 of *B. bovis* contains two identical head-totail arranged *prap-1* genes, and a single gene encoding for the RRA protein. These two loci are separated by a 88.5 kb intervening region containing ~41 genes. A comparison between the rap-1 loci of *B. bovis* and *T. equi* with the *B. microti* locus containing BmIPA48 and Bm960 genes is shown in Figure 2B. This illustration shows full synteny in the 50 and 30 ends of the *B. microti* BmIPA48 and Bm960 gene locus and the *prap-1* locus of *T. equi.* In Figure 2C we illustrate partial synteny of the 30 end of the *B. microti* genes and the rra locus of *B. bovis*. Besides the presence of the unique Cys-rich regions, there was no significant sequence similarity among the putative pRAP-1 proteins encoded by the BmIPA48 and Bm960 genes (Figure S2). However, the alignment shows conserved Cys, Tyr,

of the proteins.

and other typical residues of the pRAP-1 proteins in the NT-region of the molecules, as well as other short amino acid motifs (Figure S2). The protein encoded by gene BMR1\_ BmIPA48 also contains a series of tandem repeats in its C-terminal region, a feature that is shared with the *B. bovis* RAP-1 proteins (Figure 3 and Figure S3). Strikingly, sequence analysis of the non-coding regions immediately upstream of genes BmIPA48 and Bm960 revealed conservation of a 300-bp sequence (Figure 3 and Figure S4), suggesting that expression of these two proteins might be coordinated, despite their non-relatedness in sequence. In addtion, secondary structure sequence analysis performed in silico using TMpred suggests that BmIPA48 contains a signal peptide (aa 4–24) and a putative transmembrane (TM) region (aa 164–186) (Figure 4). Since no TM domains were previously reported in this protein, the prediction was confirmed using the alternative algorithm Phobius, which also showed the presence of a TM domain in the same region (Figure S6). Bm960 protein also contains a predicted signal peptide, two TM domains, and lack a predicted GPI anchor attachment site (Figure 4). Collectively, these features are fully consistent with expression on the surface of the parasite, as previously predicted [35]. *Pathogens* **2021**, *10*, x FOR PEER REVIEW 5 of 16 **Figure 2.** Synteny map of the *rap-1* locus of *T. equi*, putative *rap-1 B. microti*, and typical s.s. *B. bovis*. (**A**) Structure of the p*rap-1* and *rra* genes localized in the chromosome 4 of *B. bovis*. (**B**) Conserved synteny among the *rap-1* loci of *B. bovis* and *T. equi* and the BMR1\_03g00960 (BmIPA48) and BMR1\_03g00947 (Bm960) genes of *B. microti*, that encode for proteins containing the typical Cys-rich region of the pRAP-1 proteins. (**C**) Conserved synteny among the *B. bovis rra* and the *B. microti* BMR1\_03g00960 and BMR1\_03g00947 genes. *Pathogens* **2021**, *10*, x FOR PEER REVIEW 5 of 16 **Figure 2.** Synteny map of the *rap-1* locus of *T. equi*, putative *rap-1 B. microti*, and typical s.s. *B. bovis*. (**A**) Structure of the p*rap-1* and *rra* genes localized in the chromosome 4 of *B. bovis*. (**B**) Conserved synteny among the *rap-1* loci of *B. bovis* and *T. equi* and the BMR1\_03g00960 (BmIPA48) and BMR1\_03g00947 (Bm960) genes of *B. microti*, that encode for proteins containing the typical Cys-rich region of the pRAP-1 proteins. (**C**) Conserved synteny among the *B. bovis rra* and the *B. microti* BMR1\_03g00960 and BMR1\_03g00947 genes.

**Figure 3.** Schematic representation of the locus encoding for the *B. microti* RAP-1 putative proteins BmIPA48 and Bm960. A ~300 bp region upstream the two ORFs is repeated (yellow boxes). **Figure 3.** Schematic representation of the locus encoding for the *B. microti* RAP-1 putative proteins BmIPA48 and Bm960. A ~300 bp region upstream the two ORFs is repeated (yellow boxes). **Figure 3.** Schematic representation of the locus encoding for the *B. microti* RAP-1 putative proteins BmIPA48 and Bm960. A ~300 bp region upstream the two ORFs is repeated (yellow boxes).

**Figure 4.** Predicted secondary structure of the pRAP-1-like proteins BmIPA48 and Bm960 using the **Figure 4.** Predicted secondary structure of the pRAP-1-like proteins BmIPA48 and Bm960 using the Program TMpred. Predicted location of signal peptide (SP) is marked with a red bar. A dashed red line marks the boundary between predicted hydrophilic and hydrophobic transmembrane regions **Figure 4.** Predicted secondary structure of the pRAP-1-like proteins BmIPA48 and Bm960 using the Program TMpred. Predicted location of signal peptide (SP) is marked with a red bar. A dashed red line marks the boundary between predicted hydrophilic and hydrophobic transmembrane regions of the proteins.

#### Program TMpred. Predicted location of signal peptide (SP) is marked with a red bar. A dashed red *2.2. Significance of Synteny Relationships among rap-1 and rra Genes of Babesia and Theileria*

line marks the boundary between predicted hydrophilic and hydrophobic transmembrane regions of the proteins. *2.2. Significance of Synteny Relationships among rap-1 and rra Genes of Babesia and Theileria*  After identifying BmIPA48 and Bm960 as two *B. microti* encoded proteins containing non-canonical piroplasmid RAP-1 Cys-rich domains, we perfomed synteny analysis of After identifying BmIPA48 and Bm960 as two *B. microti* encoded proteins containing non-canonical piroplasmid RAP-1 Cys-rich domains, we perfomed synteny analysis of these genes. Results showed a remarkable synteny conservation of the *rap-1* locus in

these genes. Results showed a remarkable synteny conservation of the *rap-1* locus in piroplasmids (Figure 2). The BmIPA48 and Bm960 locus is in close vicinity with 3 genes that are also in the neighborhood of the *rap-1* genes in *B. bovis* and *Theileria* spp. (Figure 2). Moreover, one of these neighboring genes, encoding for the platelet-derived GF associated protein, is also associated with the locus of the *rra* gene of *B. bovis*. The schematic representation of the *rap-1* loci of *B. microti*, *Theileria* and *Babesia* s.s. in Figure 2 suggests the occurrence of a mechanism of genomic rearrangement in a chromosome of an ancestral *Babesia* organism that resulted in the insertion of an intervening region (~88

Next, we inferred on the phylogenetic relationship between amino acid sequences of

pRAP-1-like BmIPA48 and Bm960 (Clade I, *B. microti*-group: *B. microti* RI) with that of pRAP-1 proteins encoded in available reference genomes of piroplasmid species

*2.3. Phylogeny of Piroplasmid RAP-1 Proteins Recapitulates Piroplasmid Phylogeny* 

*2.2. Significance of Synteny Relationships among rap-1 and rra Genes of Babesia and Theileria* 

2). Moreover, one of these neighboring genes, encoding for the platelet-derived GF associated protein, is also associated with the locus of the *rra* gene of *B. bovis*. The schematic representation of the *rap-1* loci of *B. microti*, *Theileria* and *Babesia* s.s. in Figure 2 suggests the occurrence of a mechanism of genomic rearrangement in a chromosome of an ancestral *Babesia* organism that resulted in the insertion of an intervening region (~88

kb) encoding 41 genes in the case of *B. bovis* (Figure 2A).

Next, we inferred on the phylogenetic relationship between amino acid sequences of pRAP-1-like BmIPA48 and Bm960 (Clade I, *B. microti*-group: *B. microti* RI) with that of pRAP-1 proteins encoded in available reference genomes of piroplasmid species

*2.3. Phylogeny of Piroplasmid RAP-1 Proteins Recapitulates Piroplasmid Phylogeny* 

kb) encoding 41 genes in the case of *B. bovis* (Figure 2A).

piroplasmids (Figure 2). The BmIPA48 and Bm960 locus is in close vicinity with 3 genes that are also in the neighborhood of the *rap-1* genes in *B. bovis* and *Theileria* spp. (Figure 2). Moreover, one of these neighboring genes, encoding for the platelet-derived GF associated protein, is also associated with the locus of the *rra* gene of *B. bovis*. The schematic representation of the *rap-1* loci of *B. microti*, *Theileria* and *Babesia* s.s. in Figure 2 suggests the occurrence of a mechanism of genomic rearrangement in a chromosome of an ancestral *Babesia* organism that resulted in the insertion of an intervening region (~88 kb) encoding 41 genes in the case of *B. bovis* (Figure 2A).

#### *2.3. Phylogeny of Piroplasmid RAP-1 Proteins Recapitulates Piroplasmid Phylogeny*

Next, we inferred on the phylogenetic relationship between amino acid sequences of pRAP-1-like BmIPA48 and Bm960 (Clade I, *B. microti*-group: *B. microti* RI) with that of pRAP-1 proteins encoded in available reference genomes of piroplasmid species belonging to Clade II (Western clade: *B. duncani* WA), Clade III (Cytauxzoon: *C. felis* Winnie), Clade IV (Equus group: *T. equi* WA1), Clade V (*Theileria* s.s: *T. annulata* Ankara C9, *T. parva* Muguga, and *T. orientalis* Shintoku), and Clade VI (*Babesia* s.s.: *B. bovis* T2Bo, *B. ovata* Miyake, *B. bigemina* Bond, *Babesia* sp. Xinjiang) (Clades as defined by Schnittger et al. 2012 [1], Jalovecka et al. 2019 [3]) (Figure 5). Based on the assumption that *B. microti* is distantly related to other piroplasmid species, the tree was rooted using *B. microti* RI RAP-1-like BmIPA48 as an outgroup. As can be seen in Figure 5, the constructed pRAP-1 protein tree recapitulates phylogenetic lineages of piroplasmids as previously reported [1]. However, Bm960 places with a low bootstrap (bs: 34) as sister taxon to remaining *Babesia* s.s. pRAP-1 proteins due to its low sequence identity with other pRAP-1 proteins. *Pathogens* **2021**, *10*, x FOR PEER REVIEW 6 of 16 belonging to Clade II (Western clade: *B. duncani* WA), Clade III (Cytauxzoon: *C. felis* Winnie), Clade IV (Equus group: *T. equi* WA1), Clade V (*Theileria* s.s: *T. annulata* Ankara C9, *T. parva* Muguga, and *T. orientalis* Shintoku), and Clade VI (*Babesia* s.s.: *B. bovis* T2Bo, *B. ovata* Miyake, *B. bigemina* Bond, *Babesia* sp. Xinjiang) (Clades as defined by Schnittger et al. 2012 [1], Jalovecka et al. 2019 [3]) (Figure 5). Based on the assumption that *B. microti* is distantly related to other piroplasmid species, the tree was rooted using *B. microti* RI RAP-1-like BmIPA48 as an outgroup. As can be seen in Figure 5, the constructed pRAP-1 protein tree recapitulates phylogenetic lineages of piroplasmids as previously reported [1]. However, Bm960 places with a low bootstrap (bs: 34) as sister taxon to remaining *Babesia* s.s. pRAP-1 proteins due to its low sequence identity with other pRAP-1 proteins.

**Figure 5.** Phylogenetic neighbor joining tree inferred using amino acid sequences of pRAP-1 from the reference genomes of s.s. *Babesia* (Clade VI), s.s. *Theileria* (Clade V), *T. equi* (Clade IV), *C. felis* (Clade III), and *B. duncani* (Clade II) and the *B. microti* RAP-1 proteins BmIPA48 and Bm960 (bold fonts, red boxes). Bootstrap values of 1000 replicates are shown next to the branches. BmIPA48 is used as outgroup. The scale gives the evolutionary distance used to construct the tree. **Figure 5.** Phylogenetic neighbor joining tree inferred using amino acid sequences of pRAP-1 from the reference genomes of s.s. *Babesia* (Clade VI), s.s. *Theileria* (Clade V), *T. equi* (Clade IV), *C. felis* (Clade III), and *B. duncani* (Clade II) and the *B. microti* RAP-1 proteins BmIPA48 and Bm960 (bold fonts, red boxes). Bootstrap values of 1000 replicates are shown next to the branches. BmIPA48 is used as outgroup. The scale gives the evolutionary distance used to construct the tree.

Because of the expansion in the number of pRAP-1 domains in *Theileria* and *Cytauxzoon*, here we propose a new nomenclature for this gene family, which is based on

Interestingly, two orthologous groups of *Babesia* s.s. (*rap1d-3* and *rap1d-4*) correspond with chromosomal rearrangements that have resulted in the generation of RRA proteins, which, although their pRAP-1 domain is complete, are shortened at their C-terminal end and are only found in *Babesia* s.s. (Figure 1). As shown in Figure 2, *Babesia* parasites contain two or more copies of pRAP-1 and a single additional RRA located ~40–80 kb from the

in *B. microti*, Clade I; *B. duncani*, Clade V; and *Babesia* s.s., Clade VI), a tandemly repeated (as seen in *Cytauxzoon*, Clade III; *T. equi*, Clade IV, and *Theileria* s.s. Clade V), or tandemly triplicated pRAP-1 domains (*T. equi*, Clade IV, and *Theileria* s.s. Clade V), respectively (Table 1). Furthermore, an additional number refers to the placement into different

orthologous groups within each piroplasmid phylogenetic lineage (Figure 5).

Because of the expansion in the number of pRAP-1 domains in *Theileria* and *Cytauxzoon*, here we propose a new nomenclature for this gene family, which is based on the number of pRAP-1 domains in encoded proteins and shown in Figures 3 and 5. Thus, *rap1d*, *rap2d*, and *rap3d* genes encode for pRAP-1 proteins that comprise of a single (as seen in *B. microti*, Clade I; *B. duncani*, Clade V; and *Babesia* s.s., Clade VI), a tandemly repeated (as seen in *Cytauxzoon*, Clade III; *T. equi*, Clade IV, and *Theileria* s.s. Clade V), or tandemly triplicated pRAP-1 domains (*T. equi*, Clade IV, and *Theileria* s.s. Clade V), respectively (Table 1). Furthermore, an additional number refers to the placement into different orthologous groups within each piroplasmid phylogenetic lineage (Figure 5). *Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). **Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplas-*Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). **Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplas-*Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). *Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). *Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). *Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). *Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A). *Pathogens* **2021**, *10*, x FOR PEER REVIEW 7 of 16 pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A).


**Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplasmids. mids. mids. mids. mids. **Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplasmids. **Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplasmids. **Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplas-**Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplas-

**Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplas-

**Table 1.** Correlation of *rap* domain architecture and number of *rap-1* paralogs with phylogenetic classification of piroplas-

*2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans* A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of *2.4. BmIPA48 and Bm960 are Immunogenic during Infection in Humans*  A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming Interestingly, two orthologous groups of *Babesia* s.s. (*rap1d-3* and *rap1d-4*) correspond with chromosomal rearrangements that have resulted in the generation of RRA proteins, which, although their pRAP-1 domain is complete, are shortened at their C-terminal end and are only found in *Babesia* s.s. (Figure 1). As shown in Figure 2, *Babesia* parasites contain two or more copies of pRAP-1 and a single additional RRA located ~40–80 kb from the pRAP-1 locus, separated by the insertion of an intervening sequence (Figure 2A). Results show that this is not the case for *Theileria* parasites, which did not undergo the splitting of the *rap-1* locus due to chromosome rearrangements and thus, lack *rra* genes (Figure 2A).

#### then investigated whether sera from *B. microti* infected humans contain antibodies that then investigated whether sera from *B. microti* infected humans contain antibodies that then investigated whether sera from *B. microti* infected humans contain antibodies that BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We then investigated whether sera from *B. microti* infected humans contain antibodies that BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We then investigated whether sera from *B. microti* infected humans contain antibodies that BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We that it is a component of the secretome of the parasite. So far, the immunogenicity of BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We *2.4. BmIPA48 and Bm960 Are Immunogenic during Infection in Humans*

recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). these To tagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). then investigated whether sera from *B. microti* infected humans contain antibodies that recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). then investigated whether sera from *B. microti* infected humans contain antibodies that recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). then investigated whether sera from *B. microti* infected humans contain antibodies that recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6). A previous study identified BmIPA48 and Bm960 proteins as possible biomarkers of acute infection by using a combination of nanoparticle harvesting technology and mass spectrometry on blood derived from *B. microti* infected hamsters [37]. Even though the antigenicity of Bm960 was not investigated in detail, the protein was not recognized by global antibody screening in rodent models. Bm960 was found to be highly polymorphic among strains [35], and to be present in the plasma of infected hamsters [36], confirming that it is a component of the secretome of the parasite. So far, the immunogenicity of BmIPA48 and Bm960 proteins has remained unknown in *B. microti*-infected humans. We

then investigated whether sera from B. microti infected humans contain antibodies that recognize these two proteins. To this end, we expressed and purified recombinant HIStagged truncated forms of BmIPA48 and Bm960 proteins. The recombinant proteins were analyzed in ELISA and immunoblot using previously characterized sera from *B. microti*infected humans (Figure 6). Antibodies from four *B. microti*-infected individuals recognized BmIPA48 and Bm960 in ELISA. Immunoblot analysis showed that antibodies from infected humans reacted with a product of expected size of BmIPA48 and Bm960 recombinant proteins, as recognized by control anti-HIS monoclonal antibody (Figure 6).

**Figure 6.** Immunogenicity of BmIPA48 and Bm960 in *B. microti*-infected humans. Expression of the recombinant BmIPA48 and Bm960 containing a HIS-tag in the immunoblots was demonstrated using an anti-HIS monoclonal antibody (panels (**A**,**B**), respectively). A control lysate of cells not expressing the recombinant protein was included as a negative control (HEK 293). Immunoblots were incubated with human *B. microti* positive and negative sera. ELISAs were performed with four positive (Pos 1, Pos 2, Pos 3, and Pos4) and one negative (Neg) human serum samples were tested. A control sample incubated only with secondary anti-human IgG serum (secondary only) was also included in the ELISA analysis. \*\* *p* < 0.001. \* *p* < 0.01.

#### **3. Discussion**

In this study we identified two pRAP-1 in *B. microti*, named BmIPA48 and Bm960. Previous work indicated that BmIPA48 and Bm960 are highly expressed by *B. microti* merozoites and present in the parasite secretome [35]. The BMR1\_03g00947 protein was

previously identified erroneously as an orthologue of the *P. falciparum* gene PF3D7\_1324300, and given the designation of BmIPA48, which is kept in the present study to avoid confusion [35,36]. However, alignment of PF3D7\_1324300 and BmIPA48 reveals that their similarity is limited mainly to the glycine residues located in the tandem repeat regions of the two proteins (Figure S5). The tandem repeat region in PF3D7\_1324300 is not panconserved among *Plasmodium* proteins, suggesting that it might lack functional or structural relevance, but instead, it may work as a decoy for the immune system of the host. This observation also suggests that the repeat segment may be a result of convergent evolution, and thus BmIPA48 might not be a true orthologue of PF3D7\_1324300. Also, BmIPA48 contains non-synonymous polymorphisms, including a variable microsatellite region, that is highly antigenic and secreted, as part of tubes of vesicles during infection in mice [14,35]. In addition, electron microscopy analysis demonstrated that BmIPA48 is localized inside lipid-rich vesicles, which is consistent with their exclusive association with a membrane fraction. Also, IFA shows association of BmIPA48 with the cytoplasm of infected erythrocytes [37]. The presence of previously unnoticed TM domains in BmIPA48 is compatible with the association and export of this protein via lipid-rich vesicles to the cytoplasm of host erythrocytes and eventually to the outside of the host cell, as previously reported [14].

Considering that random gene location associations among four gene loci are highly unlikely in genomes larger than 8 Mb as those of *Babesia* and *Theileria* parasites, the data strongly suggest that the BmIPA48 and Bm960 genes are positional equivalents of the *Babesia-Theileria rap-1* genes. The biological significance of conserved gene synteny remains undefined. However, co-localization of genes may be important in epigenetic mechanisms and may influence the topology of the chromatin, which in turn can heavily influence coordinated gene expression and gene evolution. It is possible that the presence of syntenic genes results in the advantages of sharing regulatory mechanisms [38]. Sequence analysis of the non-coding regions immediately upstream of BmIPA48 and Bm960 showed conservation of a 300-bp sequence, suggesting a potential coordinated expression of these genes. Notably, a similar feature was found in other *Babesia* tandemly arranged and closely related or identical gene pairs or triplets, such as the *B. bovis rap-1s*, *msa-2s*, and *ef-1α*, that share common 50 untranslated regions [15,39–41].

A remarkable synteny conservation of the *rap-1* locus in piroplasmids is shown by the data in our study (Figure 2). In addition, the insertion of an intervening region may have resulted in the splitting of the original *rap-1* locus, favoring independent gene evolution of the two identical copies of *rap-1* and *rra* genes [15,26,42]. A similar gene organization is found in *B. bigemina* with a *rra* gene separated by a similar large intervening region from a highly diversified and complex *rap-1* locus [17]. Interestingly, the intervening region between *rra* and *rap-1* in *Babesia* parasites is located ~100 kb upstream in the same chromosome in the *B. microti*, as well as in the *T. equi* genomes. However, all the *rap-1* genes are located together in a single cluster in this group of organisms, which also lack *rra* genes. Altogether, comparative analysis of the locus encoding *B. microti* BmIPA48 and Bm960 proteins with the loci of *Babesia* and *Theileria rap-1* genes provides interesting insights on the synteny and the evolution of the genome of these parasites.

Results from the phylogenetic analysis supports the notion that Bm960 cannot be defined as pRAP-1 based on sequence identity, but only due to structural conservations and synteny. Importantly, it can also be concluded from the phylogenetic tree that pRAP-1 is a relatively complex highly polymorphic protein family that underwent multiple duplications into large gene families of paralogs, tandem duplications and triplications of the pRAP-1 cys-rich domain, and a substantial nucleotide diversification, resulting in the existence of multiple highly polymorphic pRAP-1 domains. Thus, this protein family displays a considerable complexity, typically observed for molecules that play a pivotal functional role in the parasite-host interface, such as adhesion, attachment, and invasion, or the interaction with the host immune defense [43]. We hypothesize that the generation of diversification of pRAP-1 proteins is driven by a strong positive selection to optimize adhesion and attachment to their different hosts, as is required for the evolution

of parasite host-specificity. The different copy number of pRAP-1 domains in a single protein may represent an adaptation strategy to different hosts and life cycles, enabling the parasites to invade different host species and cells. *C. felis* has two tandemly arranged pRAP-1 proteins containing canonical domains, while most *Theileria* has pRAP-1 proteins with tandemly duplicated or triplicated pRAP-1 domains. Both the *C. felis* and *Theileria* pRAP-1 domains contain 4 conserved Cys residues and a single conserved Tyr residue, as originally described in *Babesia* [15,17,26]. Considering that certain pRAP-1 features correspond with the phylogenetic classification of piroplasmid species, this finding may be exploited for the development of specific diagnostic tests. Furthermore, pRAP-1 proteins with a duplicated and/or triplicated domain architecture specify the piroplasmid lineages *Cytauxzoon*, *Theileria equi*, and *Theileria* s.s. and contrast with those that encode exclusively single-domain pRAP-1s, such as *Babesia* s.s. and *Babesia* s.l. [1].

*B. microti* contains two *prap-1*-like genes located as a single cluster in the region of the genome where s.s. *Babesia* and *Theileria* organisms contain their *prap-1* genes. These two genes have a fully conserved synteny and identical flanking genes as *Theileria* parasites, as shown in Figure 2A. This implies that the aforementioned genome rearrangements resulted in an independent evolution of RRA encoding genes, which likely occurred after *Babesia* organisms emerged as separate species from a common *Babesia* and *Theileria* ancestor. This notion is further supported by the observation that all proteins segregating into the *rap1d-3* and the *rap1d-4* orthologous groups represent RRA proteins since, although they contain a complete pRAP-1 domain, are shortened at the C-terminal end (Figure 1). This strongly suggests that the ancient RRA protein has lost its C-terminal partly due to chromosomal rearrangement and places this event before the diversification of the RRA proteins.

Considering that the antigenicity of BmIPA48 and Bm960 was not previously investigated [35–37,39], here we examined sera from *B. microti*-infected humans for the presence of antibodies against these proteins. Collectively, results of ELISA and immunoblot indicate that BmIPA48 and Bm960 are immunogenic during infection in humans, and thus should be considered for further testing as possible candidates for serological diagnosis of human babesiosis caused by *B. microti*. In addition, because of their previously established high degree of expression, surface localization, conservation, and immunogenicity, BmIPA48 and Bm960 proteins might also be promising candidates for the development of vaccines that may prevent human babesiosis.

#### **4. Materials and Methods**

#### *4.1. Expression of Recombinant B. microti pRAP-1 like Proteins*

The predicted proteins encoded by *B. microti* BMR1\_03g00947 and BMR1\_03g00960 (GenBank accession numbers: XP\_021338473 and XP\_021338474, respectively), here referred to as BmIPA48 and Bm960, respectively, were analyzed by the Kyte-Doolittle scale for the presence of hydrophobic regions as previously described [44]. As a result, 105 nt and 84 nt-long fragments located at the 50 end of BMR1\_03g00947 and BMR1\_03g00960, correspondingly, encoding hydrophobic peptide segments, were excluded from the cloning and protein expression experiments described in this work. The resulting nucleotide sequences were codon-optimized for mammalian cell expression, synthesized by GenArt Gene Synthesis (Thermo Fisher Scientific, Waltham, MA, USA) and cloned into pcDNA3.4. Recombinant plasmids containing either truncated BMR1\_03g00947 (pcDNA3.4/947) or truncated BMR1\_03g00960 (pcDNA3.4/960) were fully sequenced to confirm the presence of the target genes in frame with the cytomegalovirus promoter (data not shown). Subsequently, HEK 293 cells were transiently transfected with either pcDNA3.4/947 or pcDNA3.4/960 using polyethylenimine, as described elsewhere [45]. Expression of the recombinant truncated proteins (BmIPA48tr and Bm960tr) was confirmed by immunoblot using the anti-6xHis monoclonal antibody (clone AD1.1.10) (Bio-Rad, Hercules, CA, USA). Recombinant BmIPA48tr and Bm960tr were purified using the HisPur™ Cobalt Purification Kit following the manufacturer's protocol (Thermo Fisher Scientific). After purification, the

recombinant proteins were dialyzed using the Slide-A-Lyzer™ Dialysis cassettes (Thermo Fisher Scientific) and stored at −80 ◦C until use for ELISA and immunoblot.

#### *4.2. Human Serum Samples*

Unidentified human patient serum samples were submitted to Fuller Laboratories from Labcorp, NC for anti-*B. microti* IgG determination. No clinical data were provided for any of the specimens.

#### *4.3. ELISA Procedure*

Antigen dilution was performed by mixing 8 µL of recombinant BmIPA48 antigen (approx. 0.5 µg/µL) in 500 µL PBS buffer followed by two-fold dilutions 1:2, 1:4 and 1:8. For Bm960, 4 µL of the recombinant antigen (approx. 1 µg/µL) was mixed in 500 µL PBS buffer, and then diluted two-fold 1:2, 1:4 and 1:8. Two neighboring strips of the ELISA plate were coated with 100 µL/well of 1:4 and 1:8 antigen dilutions. The antigen-coated plates were incubated at room temperature (23–25 ◦C) overnight, then back coated by adding 100 µL/well WellChampion (Microwell Plate Blocker/Stabilizer, Kementec, Copenhagen, Denmark) to each well for 5–10 min. Plates were then decanted and allowed to dry overnight in a dark low-humidity room before use. Negative serum control was obtained from a non-reactive unidentified human patient, which tested negative in confirmatory IFA analysis. Positive controls (*n* = 4) corresponded to anti-*B. microti* IgG and IgM reactive unidentified human sera with IFA endpoint titers > 1:1024 (cat. BMG-120, Fuller Laboratories, Fullerton, CA, USA). IFA testing utilized both hamster in vivo and human type O in vitro antigen (US 10,087,412 B2 patent). All sera were diluted 1:100 in sample diluent (PBS/2 mg/mL bovine serum albumin/0.1% Tween-20). One hundred µL aliquots of diluted sera were added to ELISA plate microwells. Two rows of microwells were filled with sample diluent and were used for secondary antibody controls. Plates were covered to minimize evaporation and incubated for 60 min at room temperature. Then, plates were washed four times with wash buffer (PBS/0.1% Tween-20). One hundred µL of a working dilution of anti-human IgG (γ-chain-specific)-horseradish peroxidase (HRP) conjugate (SFG-1X, Fuller Labs) were added to each well, and the plate was covered and incubated for 30 min at RT in the dark. Microwells were washed as above and 100 µL TMB substrate was added to each well. Reactions were allowed to proceed for exactly 10 min in the dark and interrupted by adding 100 µL Stop solution (0.36 N sulfuric acid). Absorbance at 450 nm was read in a microplate reader (MultiSkan MCC/340, Titertek, Pforzheim, Germany). Absorbance values of *B. microti* positive and negative sera were compared by Student's *t*-test using Prism version 6 (GraphPad Software, San Diego, CA, USA).

#### *4.4. Immunoblot Analysis*

For human serum analysis, aliquots (45 µL) of recombinant BmIPA48 and Bm960 antigens were separated using 10% Mini-PROTEAN® TGX™ Precast Protein Gels (Bio-Rad, Cat #4561034) and transferred to PVDF membranes. Membranes were blocked with 5% milk, cut into strips, and individually incubated overnight at 4 ◦C with *B. microti*positive or negative patient sera, in a 1:250 dilution. Membranes were then washed with PBS/0.1% Tween-20 and incubated for 1 h with HRP-conjugated secondary antibody (1:10,000 dilution). Following additional washings, membranes were incubated with Opti-4CN substrate diluted in 1-part Opti-4CN diluent and 9 parts distilled water (Bio-Rad, Cat# 1708235) for 5–30 min until the desired the signal was obtained.

#### *4.5. Bioinformatic Analysis*

Secondary sequence analysis was performed using TMpred Server (vital-it.ch, accessed on 1 August 2021) and Phobius (phobius.sbc.su.se, accessed on 1 August 2021). Prediction of GPI anchor signals was carried out using PredGPI (gpcr.biocomp.unibo.it/ predgpi/, accessed on 1 August 2021). Synteny studies were carried out by exploring the Piroplasma DB database (piroplasmadb.org/piro/app, accessed on 1 August 2021).

#### *4.6. Phylogenetic Analysis*

The amino acid sequence of *B. bovis* T2Bo RAP-1 (XP\_001610908) was used in a BLASTp search, adjusting parameter settings to piroplasmid sequences (taxid:5863) and reference proteins to identify homologs in completely sequenced genomes of piroplasmid species. The genomes analyzed included *B. bigemina* strain Bond [46], *B. bovis* strain T2Bo [47], *B. ovata* strain Miyake [48], *B. microti* strain RI [35], *C. felis* strain Winnie [49], *T. annulata* strain Ankara [50], *T. equi* strain WA [51], *T. orientalis* strain Shintoku [52], and *T. parva* strain Muguga [53]. In addition, RAP-1 sequences of *B. duncani* were retrieved by courtesy from yet public unavailable genomes (*B. duncani* strain WA1: Choukri Ben Mamoun, Yale School of Medicine, New Haven, CT, USA). Finally, BmIPA48 and Bm960 were identified by delta Blast using a RAP-1 region containing 4 conserved Cys, as described before.

Altogether 34 amino acid sequences were aligned by Muscle (www.ebi.ac.uk/Tools/ msa/muscle/, accessed on 30 July 2021). In order to estimate evolutionary distances, the JTT+G (G = 5.93) was determined as best model by BIC criteria and applied [54]. After eliminating all positions with gaps and missing data, the remaining 212 positions were used for estimation of a neighbor joining tree [55]. The phylogenetic analysis was carried out using MEGA7 [56].

#### **5. Conclusions**

Findings in this study suggest that *rap-1* genes appeared early in the evolution of piroplasmid parasites, implying that expression of *prap-1* and *prap-1*-like genes is required for sustaining the life cycle of these organisms. Two tandemly arranged genes separated by an 800 bp intergenic region that includes a highly conserved putative promoter region are located in a region of the *B. microti* genome with strong synteny to the *prap-1* locus of *Babesia* and *Theileria* parasites. The organization of these two *prap-1*-like *B. microti* genes is reminiscent of the organization of the *prap-1* locus in *B. bovis* [15]. This feature, together with the presence of a single Cys-rich pRAP-1 motif in the encoded proteins resembles pRAP-1/RRA proteins of *Babesia*, rather than *Theileria*, parasites, but with identical synteny to *Theileria* parasites. The presence of a shared 300-bp region in the putative regulatory DNA regions suggests that the expression of these genes might be co-regulated. Both *B. microti* proteins contain TM domains and signal peptides, which is consistent with extracellular vesicle localization. Previous work showed that Bm960 is secreted into the sera of infected mice [35] and that BmIPA48 is strongly immunogenic in infected hamsters [14,36]. The gene structure comparison and phylogenetic analysis of the *prap-1* locus among distinct piroplasmid parasites allowed valuable insights on the genetic mechanisms involved in the evolution of the members of this piroplasmid-confined gene family. Importantly, this work also confirmed that antibodies in *B. microti*-infected humans recognized the recombinant forms of both proteins, so their potential as candidates for diagnostic assays and vaccines should be further explored.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/pathogens10111384/s1, Figure S1. Schematic representation of the structural features of RAP-1 and RRA proteins, Figure S2. Amino acid sequence alignment between the BmIPA48 and Bm960 RAP-1 like proteins, Figure S3. Amino acid sequence of the Bm947 protein, Figure S4. Schematic representation of the 300 bp region of homology among the *B. microti* RAP-1 like genes BMR1\_03g00947 (947) and BMR1\_03g00960 (960), Figure S5. Amino acid alignment between BmIPA48 (Bm) and PF3D7\_1324300 (Pf), and Figure S6. Secondary structure prediction of BmIPA48using the software Phoebius.

**Author Contributions:** Conceptualization, R.G.B. and C.E.S.; Data curation, R.G.B. and C.E.S.; Formal analysis, R.G.B., J.T., C.B.M., L.F., R.E.M., M.F.-C., L.S., H.F.A. and C.E.S.; Funding acquisition, R.G.B., C.B.M., L.F., R.E.M., M.F.-C. and C.E.S.; Investigation, R.G.B., J.T., C.B.M., L.F., R.E.M., M.F.- C., L.S., H.F.A. and C.E.S.; Methodology, R.G.B., J.T., C.B.M., L.F., M.F.-C., L.S., H.F.A. and C.E.S.; Software, R.G.B., M.F.-C., L.S. and C.E.S.; Validation, R.G.B., J.T., C.B.M., L.F., R.E.M., M.F.-C., L.S., H.F.A. and C.E.S.; Writing—original draft, R.G.B., J.T., M.F.-C., L.S. and C.E.S.; Writing—review & editing, R.G.B., J.T., C.B.M., L.F., R.E.M., M.F.-C., L.S., H.F.A. and C.E.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** RB is supported by the USDA National Institute of Food and Agriculture (NIFA) (Award Number: 2020-67015-31809; Proposal Number: 2019-05375, Accession Number: 1022541. Work supported by ARS-USDA CRIS 2090-32000-039-000-D. CBM research is supported by NIH grants AI123321, AI138139, AI152220 and AI136118, the Steven and Alexandra Cohen Foundation and Global Lyme Alliance. The financial support of MFC and LS by INTA (Instituto Nacional de Tecnologia Agropecuaria) projects 2019-PD-E5-I102-001 and 2019-PE-E5-I109-001 is acknowledged. REM is supported by ATCC's Internal Research and Development Program.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** As mentioned in the Material and Methods section, unidentified human patient serum samples used in this study were submitted to Fuller Laboratories from Labcorp, NC for anti-*B. microti* IgG determination. No clinical data were provided for any of the specimens.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors would like to acknowledge Sezayi Ozubek, Paul Lacy, Jacob Laughery, Manuel Rojas (DVM), and Jinna Navas for productive discussions and technical assistance.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Treatment of Human Babesiosis: Then and Now**

**Isaline Renard and Choukri Ben Mamoun \***

Department of Internal Medicine, Section of Infectious Diseases, Yale School of Medicine, New Haven, CT 06520, USA; isaline.renard@yale.edu

**\*** Correspondence: choukri.benmamoun@yale.edu

**Abstract:** Babesiosis is an emerging tick-borne disease caused by apicomplexan parasites of the genus *Babesia*. With its increasing incidence worldwide and the risk of human-to-human transmission through blood transfusion, babesiosis is becoming a rising public health concern. The current arsenal for the treatment of human babesiosis is limited and consists of combinations of atovaquone and azithromycin or clindamycin and quinine. These combination therapies were not designed based on biological criteria unique to *Babesia* parasites, but were rather repurposed based on their wellestablished efficacy against other apicomplexan parasites. However, these compounds are associated with mild or severe adverse events and a rapid emergence of drug resistance, thus highlighting the need for new therapeutic strategies that are specifically tailored to *Babesia* parasites. Herein, we review ongoing babesiosis therapeutic and management strategies and their limitations, and further review current efforts to develop new, effective, and safer therapies for the treatment of this disease.

**Keywords:** babesiosis; *Babesia microti*; *Babesia duncani*; parasite; therapy; atovaquone; endochin-like quinolones (ELQs)

#### **Citation:** Renard, I.; Ben Mamoun, C. Treatment of Human Babesiosis: Then and Now. *Pathogens* **2021**, *10*, 1120. https://doi.org/10.3390/pathogens 10091120

Academic Editors: Estrella Montero, Jeremy Gray, Cheryl Ann Lobo and Luis Miguel González

Received: 30 July 2021 Accepted: 27 August 2021 Published: 1 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Human babesiosis is a rapidly emerging tick-born infectious disease caused by intraerythrocytic parasites of the genus *Babesia*. Of several hundred *Babesia* species identified so far, only a few are known to infect humans. These include *Babesia microti*, *Babesia duncani*, *Babesia divergens* and *divergens*-like species, *Babesia crassa*-like, and *Babesia venatorum* [1]. In the United States, most cases of human babesiosis have been attributed to infection with *B. microti*, but sporadic cases due to infection with *B. duncani* and *B. divergens*-like MO1 have also been reported. In Europe, *B. divergens* used to be the main species responsible for infection in humans. However, recent studies suggest that *B. microti* and *B. venatorum* are now more prevalent than *B. divergens* [2]. In China, human babesiosis is mainly caused by *B. microti* and *B. venatorum*, and in the rest of the world, only a few sporadic cases have been reported and were mostly linked to *B. microti* infection [2].

*Babesia* spp. are apicomplexan parasites that infect the host red blood cells and are transmitted to mammals by tick vectors (Figure 1). The species of ticks involved in the transmission of *Babesia* pathogens vary depending on the geographical area and parasite species [1,2]. During the life cycle of *Babesia*, humans are typically accidental hosts, and most infections are linked to a tick route of transmission [1,2]. However, an increasing number of transfusion-transmitted babesiosis cases have been reported in the US over the past 2–3 decades, making *Babesia* infections a major public health concern [1,3–6]. In 2011, human babesiosis became a nationally notifiable disease in the US [5] and as one of the most common transfusion-transmitted pathogens in the US, *B. microti* was added to the list of significant threats to the blood supply [3,4]. In addition to human-to-human transmission through blood transfusion, several reports have also established the possibility of transplacental transmission from mother to child [1].

**Figure 1.** Cycle of transmission of the most common *Babesia* species, *B. microti*. During a blood meal, an infected tick introduces merozoites into the host (mouse or deer, for example). Free merozoites enter red blood cells and undergo asexual replication. While in the blood, some parasites differentiate into male and female gametocytes (not morphologically recognizable by light microscopy). These gametocytes are then taken up by a tick during a blood meal and differentiate into gametes. While in the gut, gametes fuse to form a zygote, that will subsequently undergo meiotic and several mitotic divisions to form sporozoites that are then transmitted to a mammalian host. Humans are typically accidental hosts and become infected through the bite of an infected tick. Human to human transmission is also possible via blood transfusion.

In most individuals, babesiosis remains asymptomatic or presents with mild flulike symptoms [1,2]. However, in more susceptible populations, such as the elderly, asplenic, or immunocompromised individuals, the disease can become severe and even life-threatening, with symptoms such as severe anemia, acute respiratory distress, organ failure, and death [1,2].

In the following sections, we describe the current treatment and management of *Babesia*-infected patients and their limitations. Furthermore, we report on the development and evaluation of novel and highly promising antibabesial therapies.

#### **2. Current Treatments against Human Babesiosis**

The current arsenal for the treatment of human babesiosis relies principally on four drugs: atovaquone, azithromycin, clindamycin, and quinine. Atovaquone is used to treat several human diseases, including *Pneumocystis jirovecii* pneumonia [7], toxoplasmosis [8], and malaria (in combination with proguanil (Malarone) [9]. In apicomplexan parasites, atovaquone targets the cytochrome *bc<sup>1</sup>* complex of the mitochondrial electron transport chain (Figures 2 and 3) [10–13]. Azithromycin is a relatively broad-spectrum antibiotic indicated for the treatment of numerous bacterial infections, such as those caused by *Staphylococcus* spp. [14–16] and *Legionella* spp. [17]. The antibiotic is also used for the treatment of toxoplasmosis [12] and, in combination with other drugs, for the treatment of malaria [18]. Azithromycin is a well-characterized protein synthesis inhibitor, which in apicomplexan parasites targets the translation machinery in the apicoplast (Figure 2) [19–21]. It is worth noting that azithromycin was found to have a "delayed death" effect, in which parasite division produces viable daughter cells that are subsequently unable to divide in the following cycle [19,21,22]. Clindamycin is another antibiotic commonly used for the treatment of various bacterial infections [23] and repurposed for the treatment of parasitic infections. In combination with quinine, clindamycin is used for the treatment of both malaria and babesiosis [24–26]. Several reports have suggested that clindamycin acts in a similar way as azithromycin and targets protein synthesis in the apicoplast (Figure 2) [19,21,22]. Furthermore, selection of clindamycin-resistant *T. gondii* parasites showed cross-resistance to

azithromycin, further suggesting a common target [27]. Quinine is a widely used antimalarial agent, typically administered in combination with an antibiotic such as clindamycin or doxycycline [28]. However, the drug is poorly tolerated and, as such, tends to be replaced by alternative drugs with fewer side-effects [28,29]. In malaria parasites, several modes of action for quinine have been proposed. The most commonly reported mechanism of action involves the disruption of hemozoin formation, resulting in accumulation of free ferriprotoporphyrin IX, a by-product of hemoglobin degradation, which is deleterious to parasite growth [30–32]. Confocal imaging using fluorescent derivatives of quinine and its structural analogues, quinidine and chloroquine, have shown accumulation of the probes in the digestive vacuole, consistent with the activity of this compound in this organelle [32,33]. Unlike *Plasmodium* parasites, *Babesia* species lack a digestive vacuole, do not degrade hemoglobin, and do not produce hemozoin. Therefore, the mode of action of quinine against *Babesia* parasites is likely to be different from that in *Plasmodium*. Interestingly, fluorescent probes were found to bind to phospholipids and to accumulate in membranous structures, including the parasite plasma membrane, the endoplasmic reticulum, and the mitochondrion, suggesting that quinine may inactivate specific biological functions in these organelles [32,33]. Another proposed hypothesis is that quinine acts as a DNA intercalator [34–36]. However, the lack of fluorescence in the nucleus reported by Woodland et al. seem to refute interactions with DNA as a potential mode of action [32,33]. More recently, a study in *P. falciparum* using thermal shift assays suggested that the purine nucleoside phosphorylase (PfPNP) might also be a target of quinine [37].

**Figure 2.** Schematic representation of a *Babesia*-infected red blood cell and sites of action of some approved and experimental drugs. Azithromycin and clindamycin target the apicoplast; atovaquone and ELQs target the mitochondrion. A: apicoplast, C: conoid + polar rings, DG: dense granule, ER: endoplasmic reticulum, G: Golgi apparatus, M: mitochondrion, MN: microneme, PPM: parasite plasma membrane and R: rhoptry.

**Figure 3.** Proposed mechanism of action of atovaquone and endochin-like quinolones in *Babesia* mitochondrion. (**a**) Schematic representation the mitochondrial electron transfer chain. (**b**) Schematic representation of the parasite *bc<sup>1</sup>* complex with proposed mode of action of atovaquone and ELQs.

The severity of babesiosis depends mainly on the host's immune status, the presence of risk factors and the *Babesia* species responsible for the infection. In symptomatic patients, babesiosis usually manifests with flu-like symptoms such as fever, fatigue, chills, sweats, and headache [38]. For this moderate form of the disease, typically associated with a low parasitemia level (<4%) [26], no hospital admission is required and a 7–10-day treatment course of oral atovaquone + azithromycin (500 mg azithromycin on day 1, followed by 250 mg on subsequent days + 750 mg b.i.d. atovaquone) is recommended [26,38]. Babesiosis typically resolves within seven days from the start of the treatment, but asymptomatic, low level parasitemia may persist for up to one year [26]. Monitoring of persistent parasitemia in immunocompetent individuals following treatment is usually not necessary. However, given the risk of transmission of *Babesia* parasites through blood transfusion, these patients are excluded as blood donors [3]. Immunocompromised individuals are more at risk of developing a severe form of babesiosis, resulting in complications such as acute respiratory distress syndrome, disseminated intravascular coagulation, severe hemolytic anemia, organ failure, splenic rupture, relapse, and death [2,26]. A combination of oral clindamycin + quinine (600 mg + 650 mg, every 8 h) is the standard of care for the treatment of severe babesiosis [26,38]. However, this treatment regimen is frequently associated with serious side effects, such as hearing loss, vertigo, and tinnitus. In some cases, these side effects can be so severe that dose reduction or discontinuation of treatment is required [38]. Recently, it has been demonstrated that a combination of atovaquone + azithromycin

is also suitable for the treatment of severe babesiosis, displaying comparable efficacy to clindamycin + quinine with fewer side effects [39]. Although atovaquone + azithromycin is now the preferred course of treatment for severe babesiosis, the standard 7–10-day treatment regimen of oral atovaquone + azithromycin is usually not enough to eliminate *Babesia* infection. Higher doses, longer treatment duration, and in some cases intravenous administration is required to clear the infection [26]. It is also worth noting that the use of immunosuppressive agents such as Rituximab to treat prior illnesses (B cell lymphoid malignancies, rheumatoid arthritis, etc.) may lead to babesiosis relapse and extended persistence of *Babesia* parasites [40–42].

One downside of a prolonged treatment regimen and dose escalation is the risk of developing drug resistance. Previous reports have established the emergence of mutations in the cytochrome b (Cytb) of *Babesia* parasites in humans and animal models following treatment with atovaquone [11,42,43]. In 2016, Lemieux et al. examined clinical isolates of relapsing babesiosis and identified a methionine to isoleucine mutation (M134I) in the Q<sup>o</sup> site (atovaquone-binding site) of the BmCytb [43]. This same mutation was observed in a murine model of *B. microti* infection [11], as well as in other apicomplexan parasites, such as *P. falciparum* and *T. gondii* [43]. Later, Simon et al. reported a Y272C mutation in the BmCytb Q<sup>o</sup> site in a patient presenting with relapsed *B. microti* infection following an atovaquone + azithromycin treatment course [42]. In both cases, these mutations have been shown to impact the atovaquone-binding domain [44] and appear to be associated with decreased sensitivity to the drug [42,43]. With regard to azithromycin resistance, sequencing of clinical isolates obtained from patients with relapsing babesiosis identified mutations in the ribosomal protein subunit L4 (RPL4) encoded by the apicoplast genome [42,43]. Lemieux et al. identified three substitutions in the RPL4: R86H, R86C and S73L [43]. Simon et al. observed the same R86C mutation in a patient presenting with relapsing babesiosis following atovaquone + azithromycin treatment [42]. Similar mutations associated with azithromycin resistance have been reported in *P. falciparum* [20] and *S. pneumoniae* [45] RPL4. Alternative management strategies for human babesiosis in the case of persistent relapse include the use of different drug combinations such as atovaquone + azithromycin + clindamycin, atovaquone + clindamycin, atovaquone + proguanil, or atovaquone + azithromycin + clindamycin + quinine [26,41,46,47]. The introduction of other drugs such as doxycycline, moxifloxacin, pentamidine, trimethoprimsulfamethoxazole or artemisinin to treatment regimens with the standard therapies was also reported [40,48]. A recent study in a small cohort of patients suffering from Lyme disease and babesiosis co-infection suggested improvement, and in some cases remission, following one course of disulfiram monotherapy [49]. In patients with high parasitemia (>10%), exchange transfusion is recommended and often results in a rapid reduction of the parasite load [26,50].

Despite clinical evidence that atovaquone, azithromycin, clindamycin and quinine can be used to manage human babesiosis, preclinical evaluation of these drugs in different models of *Babesia* infection has not demonstrated unanimous results with regards to their efficacy. Clindamycin showed only limited activity at a dose of 300 mg/kg (p.o.) in *B. microti*-infected Mongolian jirds [51]. When evaluated in *B. microti*-infected hamster, a course of 150 mg/kg (i.m. or p.o.) of clindamycin resulted in a two-fold decrease in peak parasitemia. Similar results were obtained when clindamycin was administered in combination with quinine [52]. AbouLaila et al. reported a ~three-fold decrease in peak parasitemia following i.p. injection of 500 mg/kg of clindamycin in *B. microti*-infected Balb/c mice [53]. Another study using the same Balb/c model of *B. microti* infection showed that oral administration of clindamycin at 25, 50, and 100 mg/kg did not lead to reduction of parasite burden [54]. Similar results were obtained by Lawres et al. following oral administration of 10 or 50 mg/kg of clindamycin to immunocomprimized mice infected with *B. microti* [11]. The consensus seems to be more apparent in the case of quinine, where most studies report no effect on parasitemia following administration of quinine as a single drug [11,52,54]. Interestingly, a combination of clindamycin + quinine was reported

to achieve up to 70% suppression of parasitemia [55] and result in a faster resolution of parasitemia compared to clindamycin alone [52], suggesting a potential synergy between the two drugs. Preclinical investigation of azithromycin efficiency against *Babesia* parasites also yielded inconsistent results. In *B. microti*-infected Balb/c mice, a four-day treatment course with azithromycin at 25, 50, and 100 mg/kg was found to be potent, resulting in 75–96% suppression of parasitemia [54]. In contrast, the evaluation of azithromycin in *B. microti*-infected SCID mice showed no effect on parasitemia at 10 and 50 mg/kg after a seven-day treatment course [11]. Similar results were obtained in *B. microti*-infected hamsters, where 150 mg/kg azithromycin treatment regimen, administered daily for almost two weeks, showed no apparent effect on parasitemia [56]. Out of the four clinically used drugs in the treatment of babesiosis, only atovaquone seems to consistently show high potency against *Babesia* parasites [11,56–59]. Studies carried out in *B. microti*-infected hamsters and SCID mice reported fast clearance of parasitemia following treatment with atovaquone [11,56]. However, recrudescence due to atovaquone-resistant parasites was observed [11,56]. In *B. microti*-infected hamsters, a combination therapy of atovaquone + azithromycin resulted in rapid clearance of parasitemia without recrudescence [56]. In a lethal model of *B. microti* infection in hamsters, atovaquone monotherapy was found to be superior to a combination of clindamycin + quinine, resulting in low to undetectable parasitemia and extended survival [58]. Potency of atovaquone was also demonstrated in *B. divergens* [59] and *B. duncani* [57] models, with IC<sup>50</sup> values in the low nanomolar range. In gerbils, although prophylaxis experiments were not successful, a dose of atovaquone as low as 0.5 mg/kg was found to efficiently prevent *B. divergens* infection, so long as daily treatment was maintained several days post-infection [59]. In the case of *B. duncani*, a treatment course of 10 mg/kg atovaquone resulted in a clear reduction of parasitemia and 80% survival using a mouse model of lethal infection [57]. The results derived from the evaluation of atovaquone, azithromycin, clindamycin, and quinine in preclinical models of babesiosis are summarized in Table 1.

While combinations of atovaquone + azithromycin and clindamycin + quinine have been used for more than 20 years for the treatment of human babesiosis [60], the efficacy of these drugs and their primary modes of action in *Babesia* parasites have only recently started to be elucidated.


**Table 1.** Reported efficacy of atovaquone, azithromycin, clindamycin and quinine in animal models of babesiosis.


**Table 1.** *Cont.*

#### **3. In Vitro and In Vivo Models for the Evaluation of Novel Anti-***Babesia* **Therapies**

Evaluation of the potency of novel drugs for the treatment of human babesiosis has proven challenging due to the absence of a continuous in vitro culture system for *B. microti*, the main causative agent of human babesiosis. A *B. microti* short-term ex vivo system has been used previously for growth inhibition assays [11,62]. However, this culture system is not amenable for high-throughput screening of large libraries of compounds. Despite the current challenges faced in the development of a stable *B. microti* in vitro culture system, this parasite can easily be propagated in rodents, such as mice [63–65], hamsters [66], and gerbils [51]. Two very distinct profiles of *B. microti* infection in preclinical models have been observed, depending on the immune status of the host. In immunocompetent animals, such as Balb/c mice, golden hamsters, or gerbils, the parasitemia typically rises within

a few days following infection, reaches a peak (40–60% parasitemia), and then resolves on its own [63–65]. In immunocompromised animals, such as SCID and rag2D mice, the parasitemia rises and then plateaus at ~50–80% parasitemia [11,57,63]. Immunocompromized mice infected with *B. microti* maintain high parasitemia levels over time but do not succumb to infection [11,57,63]. Although the most commonly used *B. microti* preclinical models (described above) are non-lethal, one research group reported the use of a lethal model of *B. microti* infection in hamsters using the ATCC30222 strain [58,67]. In this model, parasite inoculation results in fulminating disease reaching 90% parasitemia and almost 100% mortality by DPI 12 [58,67]. This model of infection was previously used to evaluate the potency of atovaquone [58].

In vitro culture of *B. divergens*, a species known to infect humans and cattle [1], has been established in mammalian erythrocytes and can be used for the evaluation of potential drug candidates [59,68–73]. An in vivo model of *B. divergens* is available in gerbils [74,75] and has been used for the evaluation of potential antibabesial drugs [59,76]. Multiple other rodent species such as rats, mice, hamsters or guinea pigs were tested for the establishment of infection, but none developed parasitemia [74].

The first in-vitro culture system of *B. duncani* in hamster red blood cells was established in 1994 [77]. More recently an adapted protocol of *B. duncani* culture in hamster RBCs was reported using another culture medium [78]. The authors also investigated alternate RBC sources such as mouse, rat, horse or cow. None of these RBCs were able to sustain *B. duncani* growth [78]. In 2018, Abraham et al. reported the first continuous in vitro culture system for *B. duncani* in human erythrocytes [79]. The development of this system allowed for the high-throughput screening of novel derivatives for the treatment of human babesiosis [57]. *B. duncani* can be propagated in hamsters and typically results in fatal infection following the development of pulmonary edema and respiratory distress [80,81]. However, to the best of our knowledge, the *B. duncani* hamster model was not used for the assessment of potential antibabesial drugs. *B. duncani* infection can also be established in mice and is associated with a fatal outcome in specific mouse genetic backgrounds [82,83]. Similar to the hamster model, *B. duncani*-infected mice present with pulmonary edema, leading to respiratory distress and death [83]. Interestingly, it was shown that susceptibility to acute babesiosis following *B. duncani* infection is significantly influenced by the gender and genetic background of the animal [82]. Recently, Chiu et al. presented the first use of a lethal model of *B. duncani* infection in mice for the evaluation of novel promising candidates for the treatment of human babesiosis [57]. The different models of in vitro and in vivo *B. microti*, *B. divergens,* and *B. duncani* available for the evaluation of novel therapeutics are summarized in Table 2.



Overall, there is a wide variety of *Babesia* models available. However, finding complementary systems can prove challenging. Even though *B. microti* accounts for the majority of human babesiosis cases, the absence of a continuous in vitro culture system makes it challenging to use this species for drug discovery purposes. On the other hand, *B. divergens* can be used for in vitro drug screening. However, it's in vivo model using gerbils may not be widely accessible. Considering this, *B. duncani* appears as the *Babesia* species of choice for drug development. The availability of a stable in vitro culture system in human red blood cells allows for high-throughput screening of large libraries of candidates, offering the possibility to conduct detailed structure–activity relationship studies. Furthermore, the

availability of a reproducible model of *B. duncani* lethal infection in immunocompetent mice offers a reasonably affordable option to assess promising drug candidates.

#### **4. Novel Therapies under Investigation for the Treatment of Human Babesiosis**

The recent effort to develop new therapeutics for the treatment of human babesiosis has mostly focused on repurposing known anti-piroplasm agents. A large library of antimalarial drugs, such as artesunate, artemether, dihydroartemisinin, chloroquine, mefloquine, piperaquine, halofantrine, lumefantrine, pyrimethamine, and pyronaridine, has been assessed against *B. microti* but failed to demonstrate much, if any, efficacy against parasite load at the selected dose [51,54,55]. Other antimalarials such as primaquine, pentaquine and robenidine showed potent parasitemia suppression in *B. microti*-infected animals [51,54]. Screening of the Malaria Box, a 400-compound library with known antimalarial activity [84], led to the identification of nine compounds with low micromolar/nanomolar potency (2.1 µM to 160 nM) against *B. divergens* cultured in human erythrocytes [71]. To the best of our knowledge, no further evaluation of the most promising candidates has been reported so far. More recent screenings of the Malaria Box and the Pathogen Box (400 compounds active against neglected diseases) also reported 38 and nine compounds, respectively, with nanomolar potency against *Babesia* species responsible for bovine (*B. bovis* and *B. bigemina*) and equine (*B. caballi*) babesiosis [85,86]. The two most promising compounds identified from the Pathogen Box screening were further assessed in *B. microti*-infected Balb/c mice and showed significant reduction of peak parasitemia [85].

Over the recent years, a large number of drugs, including actinonin [87], atranorin [61], N-acetyl-L-cystein [88], chalcone-4-hydrate [89], trans-chalcone [89], cryptolepine [90], ellagic acid [91], eflornithine [92], fusidic acid [93], gossypol [94], gedunin [95], hydroxyurea [92], luteolin [87,95], nimbolide [95], pepstatin A [96], xanthohumol [94], fluoroquinolone derivatives (enrofloxacin, enoxacin, norfloxacin, ofloxacin, trovafloxacin) [97,98], ciprofloxacin and some of its novel derivatives [53,99], and natural extracts of *Syzygium aromaticum* [100], *Camellia sinensis* [100], *Cinnamomum verum* [101], *Olea europaea,* and *Acacia laeta* [102] have been assessed for antibabesial properties. The in vitro evaluation of these derivatives was mainly carried out against the species responsible for bovine and equine babesiosis and revealed, in most cases, growth inhibition in the micromolar range. In vivo evaluation of these compounds was typically performed in *B. microti*infected hamsters or Balb/c mice. In most cases, diminution and/or delay in peak parasitemia was observed, but none of the monotherapies displayed high potency against *B. microti* [61,90–92,94,97,100,101]. Although some of these compounds could turn out to be promising for veterinary use, they are unlikely to be accepted for clinical use based on their poor selectivity indices. Despite this fact, some of these drugs could be investigated as starting points for structural optimization for the development of novel antibabesial agents.

Out of the multiple derivatives recently reported with potent anti-*Babesia* efficacy, tafenoquine, clofazimine, and endochin-like quinolones are probably the most promising drugs.

Tafenoquine, previously known as WR238605, is an 8-aminoquinoline. In 2018, tafenoquine was approved by FDA for the radical cure of *Plasmodium vivax* infection and for chemoprophylaxis of malaria [103]. Several research groups have investigated the potential of tafenoquine against *Babesia* parasites [55,104–106]. In 1997, Marley et al. reported that a twice daily injection of Tafenoquine (i.m.) at 52 mg/kg for four days resulted in 100% parasitemia suppression by day 3 post-drug removal in *B. microti*-infected golden hamsters [55]. Furthermore, a subpassage experiment was carried out to determine whether parasitologic cure was achieved. None of the animals that received blood from tafenoquinetreated hamsters became parasitemic after six weeks post-administration, indicating that treatment resulted in complete cure of *B. microti* infection [55]. More recently, Mordue et al. evaluated tafenoquine in *B. microti*-infected SCID mice [106]. When parasitemia reached ~10%, mice were administered with a single dose of 20 mg/kg of tafenoquine (p.o.). By day 4 post-treatment, parasitemia level was undetectable in tafenoquine-treated animals and

remained so until the end of the experiment (day 18 post-treatment). To assess whether the lack of detection of parasites in blood smears was indicative of cure, blood collected from tafenoquine-treated animals at day 18 PT was injected in naive SCID mice. The newly inoculated animals developed detectable parasitemia within one week. Interestingly administration of a single dose of 20 mg/kg tafenoquine (p.o.) resulted in undetectable parasitemia within four days, indicating that the parasites remained sensitive to tafenoquine. In the case where mice were kept beyond day 18 PT, recrudescence was observed day 37 PT. In a separate experiment, *B. microti*-infected mice were treated with a first dose of 25 mg/kg of tafenoquine (p.o.) when parasitemia reached ~20%, followed by a second dose of 12.5 mg/kg of tafenoquine (p.o.) three days later to account for the decrease of plasma concentration. By day 5 after administration of the first dose, parasitemia was below detection level and remained so until day 28 PT. Subpassage of blood collected at day 28 PT in naïve SCID mice resulted in detectable parasitemia with nine days post-inoculation. Overall, although no radical cure was achieved in these experiments, a single oral dose of tafenoquine was found efficient to rapidly reduce parasitemia burden. It is also worth noting that despite the recrudescence observed following treatment, re-emerging parasites did not develop resistance to tafenoquine and remained susceptible to the drug [106]. In 2020, Carvalho et al. investigated tafenoquine in *B. microti*-infected Balb/c mice. Potent inhibition was observed following administration of 10 mg/kg of tafenoquine (three doses on alternate days, p.o.) or of a combination of 10 mg/kg of tafenoquine (three doses on alternate days, p.o.) + 25 mg/kg artesunate (five daily doses, i.p.), starting at day 4 post-infection [104]. In both cases, a ~5.6-fold reduction in peak parasitemia was observed and parasitemia was undetectable from DPI 9 by examination of Giemsa-stained thin-blood smears and remained so until the end of the study (DPI 30). However, except for one animal from the tafenoquine + artesunate treatment group, all mice remained positive for *B. microti* infection by PCR at DPI 27. Interestingly, subpassage of blood collected from tafenoquine-treated mice in a naïve Balb/c mouse resulted in the development of parasitemia, whereas the mouse receiving blood from combination-treated animals remained negative [104].

Based on the results described above (summarized in Table 3), tafenoquine could be an interesting drug candidate for further evaluation for the treatment of human babesiosis. With its extended half-life in humans (12–17 days) [106], only a few doses may be required, thus limiting the development of drug resistance. One downside, however, is that tafenoquine causes severe hemolytic anemia in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency [105,107,108], and as a result its use is contraindicated in such cases. While the exact mechanism of action of tafenoquine in *Babesia* parasites remains unknown, one hypothesis is that the 8-aminoquinoline mediates oxidative stress within the parasite without damaging the host red blood cells of individuals with active G6PD [105]. The latter enzyme plays a key role in the production of NADPH and protects red blood cells from damage by reactive oxygen species (ROS). In the case of G6PD deficiency, NADPH is at a level that is not enough to protect the RBCs from tafenoquine-induced oxidative stress [105].

**Table 3.** Preclinical evaluation of promising new therapeutics for the treatment of human babesiosis: tafenoquine, clofazimine and endochin-like quinolones (ELQs).



**Table 3.** *Cont.*

Clofazimine is an antibiotic used to treat leprosy [111] and drug-resistant tuberculosis [112]. In 2016, Tuvshintulga et al. reported that clofazimine has potent antibabesial effect, following its evaluation in *B. microti*-infected Balb/c mice. A five-day treatment course of 20 mg/kg clofazimine administered either i.p. or p.o. led to suppression of parasitemia by more than 80%, with a slightly superior efficacy when administered orally [113]. Interestingly, although no parasites could be detected by blood smears, blood, heart, spleen,

kidney, and liver samples obtained from clofazimine-treated animals tested positive for the presence of *B. microti* ss-rRNA at DPI 40. Consistently, subpassage of blood collected from clofazimine-treated animals in naïve mice resulted in reinfection [113]. It is worth noting that no toxicity was observed in mice during treatment administration. Furthermore, daily administration of 200–300 mg for >30 months for the treatment drug-resistant tuberculosis in humans was well tolerated [114]. More recently, the same research group reported on the efficacy of clofazimine in *B. microti*-infected SCID mice [110]. The continuous administration of clofazimine from DPI 4 to 57 at a daily dose of 20 mg/kg resulted in undetectable parasitemia by examination of blood smears from DPI 14 onward. Parasite DNA could no longer be detected by PCR from DPI 54 until the end of the study (DPI 90), suggesting that this seven-week treatment course is efficient in curing *B. microti* infection [110]. *B. microti*infected SCID mice treated with 20 mg/kg clofazimine for seven days (DPI4-10) showed no parasitemia by DPI 24. However, recrudescence was observed from DPI 26. Initiation of a second treatment course of clofazimine failed to clear parasitemia, suggesting that the rise of recrudescent parasites was associated with development of clofazimine resistance. Blood samples obtained from two mice that developed recrudescence were sub-passaged in naïve Balb/c mice, which subsequently underwent a five-day clofazimine treatment course. Interestingly, clofazimine successfully impacted the rise of parasitemia in one case, but not in the other [110]. Sequencing analysis established that, unlike atovaquone, clofazimine does not target the cytochrome b of the parasite. As a result, atovaquone-resistant parasites were generated in SCID mice and then propagated in Balb/c mice. A two-week course of 20 mg/kg clofazimine successfully cleared infection in all the mice. However, a relapse was observed in some of the animals, which responded to a second two-week course of a higher dose of clofazimine (40 mg/kg) [110]. Based on these results, clofazimine appears as a promising candidate for the treatment of human babesiosis. Due to the risk of development of drug resistance with a short-term monotherapy, it would be interesting to evaluate the efficacy of clofazimine when combined with a partner drug such as atovaquone. Results derived from the preclinical evaluation of clofazimine are summarized in Table 3.

A novel class of compounds, endochin-like quinolones (ELQs) has recently been reported with high potency against *B. microti* and *B. duncani* [11,57]. Previously reported for their high potency against other apicomplexan parasites such as *Plasmodium* [115–125], *Toxoplasma* [126,127] and *Leishmania* [128], ELQs have been shown to target the cytochrome *bc<sup>1</sup>* complex of the parasites (Figure 3) [117,120,122,128–130]. In 2016, Lawres et al. demonstrated potency of ELQ-271 and ELQ-316 in the short-term ex vivo culture system of *B. microti* as well as in the in vivo SCID model of *B. microti* infection. In *B. microti*-infected mice, a seven-day oral administration of 10 mg/kg of ELQ-271 or ELQ-316 resulted in clearance of parasitemia, followed by recrudescence by day 12 post-drug removal [11]. Due to the high crystallinity and low aqueous solubility of this class of compounds, which precludes administration of higher doses, a prodrug of ELQ-316, ELQ-334, was designed by esterification of the carbonyl group present in the quinolone core of the molecule [11,115,116,127]. This strategy led to improved aqueous solubility and increased plasma concentration of the drug following administration of molar equivalents [11,115,116,127]. Administration of ELQ-334 as a monotherapy at 10 mg/kg in *B. microti*-infected mice resulted to slightly extended clearance of parasitemia compared to treatment with ELQ-271 and ELQ-316. However, re-emerging parasitemia was observed by day 16 post-drug removal. In all cases, recrudescence was accompanied by a GCT → GTT mutation in the Q<sup>i</sup> site of the parasite's cytochrome *bc<sup>1</sup>* complex, resulting in an Ala to Val substitution at codon 218 [11]. Since monotherapy is not the ideal treatment regimen, a combination of ELQ-334 + atovaquone was evaluated and resulted in complete clearance of parasitemia with no recrudescence following administration of doses as low as 5 + 5 mg/kg [11]. More recently, Chiu et al. reported the screening of a new library of ELQ derivatives against *B. duncani* and identified three potent ELQ prodrugs: ELQ-331 (IC<sup>50</sup> = 141 ± 22 nM), ELQ- 468 (IC<sup>50</sup> = 15 ± 1 nM), and ELQ-502 (IC<sup>50</sup> = 6 ± 2 nM). The previously reported ELQ-316 and its prodrug, ELQ-

334 were also assessed against *B. duncani* and showed IC<sup>50</sup> values of 136 ± 1 nM and 193 ± 66 nM, respectively [57].

Further evaluation of the lead candidate, ELQ-502, showed low toxicity in mammalian cells, and thus a highly desirable therapeutic index (>833). ELQ-502 was assessed in *B. duncani*- and *B. microti*-infected mice as a single drug (10 mg/kg) and in combination with atovaquone (10 + 10 mg/kg). Following a 10-day treatment course, both the mono- and the combination therapies resulted in radical cure with no recrudescence, and in the case of *B. duncani*-infected mice, 100% survival [57]. Interestingly, a shorter treatment duration with ELQ-502 alone at 10 mg/kg in *B. microti*-infected mice resulted in recrudescence [109]. Similarly to the results obtained following treatment with ELQ-271, ELQ-316, and ELQ-334, recrudescence following ELQ-502 shorter treatment duration was associated with GCT → GTT mutation in the Q<sup>i</sup> site of the *BmCytb* [109]. Results obtained from the evaluation of ELQ derivatives are summarized in Table 3.

#### **5. Conclusions and Considerations for Future Drug Development**

Human babesiosis is an emerging tick-borne disease of rising incidence and a major public health concern. The current therapies for the treatment of human babesiosis are based on drugs already in use against other apicomplexan parasites and tend to be associated with significant adverse effects and/or the development of drug resistance. Moreover, the evaluation of these drugs, namely atovaquone, azithromycin, clindamycin, and quinine, in animal models of babesiosis has raised concerned about their efficacy in achieving parasite elimination. In light of these findings, the need for novel treatments specifically designed to tackle *Babesia* infection becomes apparent. Over the past decades, there has been a growing effort to develop such therapies. Based on their potency, selectivity, and ability to eliminate infection with no recrudescence when combined with atovaquone, endochin-like quinolones (ELQs) appear to be the most promising candidates to advance the treatment of human babesiosis. With regard to the identification of novel molecules with potency against human babesiosis, it could be interesting to establish a standardized protocol for the evaluation of new candidates, in order to facilitate a comparison of results between different research centers. A consensus protocol agreed upon by members of the community and one that follows standard methods for efficacy and safety using established in vitro cell culture assays and in vivo mouse models is warranted.

**Author Contributions:** Conceptualization, C.B.M. and I.R.; writing-original draft preparation, I.R; writing-review and editing, I.R. and C.B.M.; supervision, C.B.M.; funding acquisition, C.B.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** CBM research is supported by NIH grants (AI-123321, AI-138139, AI-152220, AI-153100, AI136138 and the Steven & Alexandra Cohen Foundation and the Global Lyme Alliance).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** No new data were created. The information presented in this review article are from published reports available in public databases.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

