Stable transformation of Babesia bigemina and Babesia bovis using a single transfection plasmid

Babesia bigemina and Babesia bovis, are the two major causes of bovine babesiosis, a global neglected disease in need of improved methods of control. Here, we describe a shared method for the stable transfection of these two parasites using electroporation and blasticidin/blasticidin deaminase as a selectable marker. Stably transfected B. bigemina and B. bovis were obtained using a common transfection plasmid targeting the enhanced green fluorescent protein-BSD (egfp-bsd) fusion gene into the elongation factor-1α (ef-1α) locus of B. bigemina and B. bovis under the control of the B. bigemina ef-1α promoter. Sequencing, Southern blotting, immunoblotting and immunofluorescence analysis of parasite-infected red blood cells, demonstrated that the egfp-bsd gene was expressed and stably integrated solely into the ef-1α locus of both, B. bigemina and B. bovis. Interestingly, heterologous B. bigemina ef-1α sequences were able to drive integration into the B. bovis genome by homologous recombination, and the stably integrated B. bigemina ef-1α-A promoter is fully functional in B. bovis. Collectively, the data provides a new tool for genetic analysis of these parasites, and suggests that the development of vaccine platform delivery systems based on transfected B. bovis and B. bigemina parasites using homologous and heterologous promoters is feasible.

Together, these observations suggests that it would be feasible to use a common strategy for gene integration in both B. bovis and B. bigemina, based on targeting one of the two identical ef-1α open reading frames (orfs) present in the ef-1α locus.
Here, we describe for the first time a stable transfection system for B. bigemina based on integration of the egfp-bsd gene under the control of the ef-1α promoter, into the ef-1α locus of B. bigemina, after drug selection with blasticidin. In addition, the same plasmid used for transfection of B. bigemina was able to insert and express foreign sequences in the ef-1α locus of B. bovis, thus expanding the options available for the genetic manipulation of Babesia sp. parasites more generally.

Materials and Methods
Ethics statement. Animals  In vitro parasite culture. B. bigemina (Puerto Rico strain) 11 and B. bovis (T 3 Bo strain) 12 were propagated in continuous microaerophilic stationary-phase culture as previously described. Briefly, B. bigemina and B. bovis cultures were grown in 96-well plates in bovine red blood cells (RBC) at 5% or 10% hematocrit, respectively, using HL-1 culture media at pH 7.2, and were incubated at 37 °C in an atmosphere of 5% CO 2 and 90% N 2 10 . Evaluation of sensitivity of B. bigemina and B. bovis to blasticidin. B. bigemina and B. bovis parasites were cultured in 180 µl of culture medium containing 5% and 10% bovine RBC, respectively, in a 96-well plate with different concentrations of blasticidin: 1, 1.2, 1.5, 1.8, 2, 4, 6, 8, 10, 12 and 14 μg/μl. Media without blasticidin was used as a control. The initial parasitemia was 0.2%. One hundred and fifty µl of culture media was daily replaced with corresponding amount of blasticidin. Percentage of parasitized erythrocytes (PPE) was monitored daily by Diff-Quik stained smears every 24 hr over a period of 72 hr by light microscopy (1000 × magnification). This experiment was carried out in triplicate.
Parasite DNA extraction. Genomic DNA (gDNA) and plasmid DNA (pDNA) was extracted from cultured B. bigemina and B. bovis using the Qiagen Blood core kit according to the manufacturer's instructions.
Plasmid constructs. The 3′ and 5′ insertion target regions for stable transfection in the ef-1-alpha-B orf were generated by PCR with B. bigemina gDNA isolated from tissue culture cells utilizing the following sets of primers: Bbig-EF-orf-B-3′-BamHI-F and Bbig-EF-orf-B-3′-BamHI-R primers for the 3′ region of B. bigemina ef (expected size of 678 base pair (bp) amplicon]; Bbig-EF-orf-B-5′-Xho-F and Bbig-EF-orf-B-5′-Xho-R for the 5′ region of B. bigemina ef (expected size of 674 bp amplicon). These PCR products were cloned into the TOPO-TA 2.1 (Life technologies) cloning vector for sequence confirmation. The 3′ and 5′ insertion regions were then digested from the cloning vectors respectively with BamHI and XhoI restriction enzyme and the amplicons were isolated and recovered from a 1% agarose gel. The plasmid containing the ef-1α-A promoter and luciferase gene along with the 3′ rap-1a stop region used in the B. bigemina transient transfection 10 was digested with EcoRI restriction enzyme to remove the luciferase gene and the resulting linearized vector now containing just the B. bigemina ef-1a-A promoter and B. bigemina 3′ rap-1a stop region was re-circularized by ligation and transformed into TOP-10 E. coli competent cells (Life Technologies). The purified vector was then digested with BamHI restriction enzyme and the 674 bp amplicon corresponding to 3′ ef-1α-B orf insertion region was ligated into this linear vector, transformed into TOP-10 cells and sequenced to confirm the correction orientation of the insertion. The resulting vector was digested with XhoI restriction enzyme and a similar procedure was used to ligate the 5′ ef-1α-B orf insertion target region into the vector. After confirming that both the 3′ and 5′ insertion sites were present in the correct orientation by sequencing, the vector was prepared for further ligation by digestion with EcoRI restriction enzyme. A synthetic egfp-bsd fusion gene from the p6-Cys-EKO plasmid (GenBank Accession number KX247384) 13 containing EcoRI restriction sites was digested with EcoRI to remove the egfp-bsd fragment. This egfp-bsd fragment was isolated on a gel and the recovered fragment ligated into the vector containing the 3′ and 5′ B. bigemina insertion regions and the rap-1a 3′ stop region to generate the final stable transfection vector designated pbig-ef-egfp-bsd. The plasmid pbig-ef-egfp-bsd was used to transform TOP-10 cells, and plasmid was purified with the Qiagen Endotoxin free Plasmid Maxi Kit following the manufacturer's instructions prior to transfection. The pBluescript (pBS) plasmid was used as a negative control in the transfection experiments.

Transfection of B. bigemina. The transfection of B. bigemina-infected RBC was performed as described by
Suarez et al. 2 . Briefly, B. bigemina iRBC with PPE ~20% was centrifuged at 600 xg for 5 min, and the cells washed once with 1 ml of Cytomix buffer (120 mM KCl, 0.15 mM CaCl 2 , 10 mM K 2 HPO 4 /KH 2 PO 4 pH 7.6). Twenty μg of pbig-ef-egfp-bsd plasmid in 55 μl of cytomix was then gently mixed with 40 μl of washed B. bigemina iRBC then transferred to 0.2 cm electroporation cuvette and transfected by electroporation using a BioRad Gene Pulser II system at 1.2 kV + 25 µF + 200 Ω. After transfection, cells were immediately transferred to a 24-well plate containing 1.2 ml of HL-1 media with 5% of bovine RBC and 8 μg/μl of bsd to select for egfp-bsd expressing transgenic parasites.

Transfection of B. bovis.
The process for transfection of B. bovis parasites using plasmid pbig-ef-egfp-bsd was performed as described above for B. bigemina. After transfection, cells were immediately transferred to a 24-well plate containing 1.2 ml of HL-1 media with 10% of bovine RBC and 3 μg/μl of bsd to select for egfp-bsd expressing transgenic parasites. and B. bigemina wild-type parasites were cultured in 180 µl of culture medium containing 5% bovine RBC, in the absence of blasticidin and in the presence of 8 μg/μl of blasticidin. B. bovis (bovis-big-ef-egfp-bsd) and B. bovis wild-type parasites were cultured in 180 µl of culture medium containing 10% bovine RBC in the presence of 3 μg/μl of blasticidin or in the absence of blasticidin (positive control). The initial parasitemia was 0.2% and parasites were cultured in triplicate in 96-well plates. Medium (150 µl) was replaced daily. PPE was monitored every 24 hr up to 72 hr by Diff-Quik-stained RBC smears by optical microscope under 1000 × amplification. This experiment was carried out in triplicate.
PCR amplification and sequencing. Sets of primers were designed to confirm the integration of plasmid bigemina-big-ef-egfp-bsd and bovis-big-ef-egfp-bsd by specific amplification of a DNA fragments surrounding the 5′ recombination site, the 3′ recombination site and the locus (Table 1). Amplicons were cloned into pCR ™ 2.1-TOPO vector (Life Technologies) according to the manufacturer's instructions and nucleotide sequences were confirmed by Sanger sequencing (ABI 3730).
Statistical analysis. Statistical significance was determined using ANOVA (GraphPad Prism 7 software). P < 0.05 were considered statistically significant.

Growth inhibitory concentrations of blasticidin in B. bigemina and B. bovis. First we compared
the inhibitory concentrations of blasticidin on in vitro cultured B. bigemina and B. bovis. Parasites were cultured in the presence of different concentrations of blasticidin ranging from 1 µg/µl to14 µg/µl, and the PPE was calculated daily up to 72 hr (Suppl. Figure 1a and b). The results show that increasing blasticidin concentrations above 4 μg/μl correlated with decreasing PPE for B. bigemina. The calculated IC 50 found for B. bigemina and B. bovis was 3 µg/µl and 0.8 µg/µl, respectively. However 8 µg/µl and 3 µg/µl of blasticidin completely inhibited the growth of B. bigemina and B. bovis respectively (Suppl. Figure 1a and (Fig. 2a). As expected, no eGFP fluorescence was observed in B. bigemina and B. bovis wild-type control parasites (Fig. 2b).

Phenotypic comparison between non-transfected and transfected Babesia parasites. The in
vitro growth rate of bigemina-big-ef-egfp-bsd and bovis-big-ef-egfp-bsd parasite lines, and wild-type parasites were compared. Notably, bigemina-big-ef-egfp-bsd parasites grew three times faster than wild-type parasites (P < 0.05) while bigemina-big-ef-egfp-bsd parasites grew at similar rate regardless of the presence or absence of blasticidin (Fig. 3a). However, as expected, B. bigemina wild-type parasites did not grow in the presence of blasticidin (Fig. 3a).
The bovis-big-ef-egfp-bsd parasites showed a higher rate of growth (P < 0.05) in blasticidin-free culture media (Fig. 3b) when compared to growth of bovis-big-ef-egfp-bsd parasites in the presence of blasticidin or the growth of wild-type parasites. Growth rates for big-ef-egfp-bsd and wild-type parasites in the presence or absence of blasticidin were indistinguishable (Fig. 3b). B. bovis wild-type parasites did not grow in the presence of blasticidin (Fig. 3b).
Genotypic and proteomic characterization of stable B. bigemina and B. bovis transfected parasites. The genotypic characterization of transfected parasites was performed by Southern blot analysis, PCR and sequencing of PCR products. Genomic DNA from transfected bigemina-big-gfp-bsd and bovis-big-gfp-bsd cell lines, non-transfected B. bovis and B. bigemina parental strains, and plasmid pbig-ef-egfp-bsd, were analyzed in Southern blots using B. bovis msa-1 (Fig. 4a), B. bigemina rap-1 (Fig. 4b), egfp (Fig. 4c) and ef-1α (Fig. 4d) specific dig-labeled probes. The presence of a single band hybridizing with egfp-bsd probe, only in the transfected parasites is consistent with a single site integration of the exogenous transfected egfp-bsd gene in both bigemina-big-ef-egfp-bsd and bovis-big-ef-egfp-bsd cell lines (green boxes, Fig. 4c). Probing the blots with an ef-1α specific probe that hybridizes with sequences that are not included in the transfection constructs reveal an increase in the size of the ef-1α locus in the transfected parasites as a result of the insertion of the transfected genes. The calculated size of the BgIII restriction size containing the ef-1α locus of wild type B. bigemina and B. bovis are 16.4 and 18.8 Kb, respectively. However, the size of the locus, as detected by the specific labeled probe, was increased to 18.6 and 21 Kb, respectively, in both transfected parasite lines which matches with the predicted size of the stably inserted DNA (2.2 kb). As expected, the rap-1 and msa-1 probes react with identical patterns in transfected and non-transfected gDNA. Taken together, these results are consistent with stable integration of a single egfp-bsd gene copy in both parasite species.
We then designed a PCR aimed at demonstrating correct integration of the exogenous gene into the B. bigemina ef-1α locus using primers based on the transfected egfp gene and sequences adjacent to the ef-1α locus of B. bigemina and B. bovis that are not present in the transfection plasmids (Figs 5 and 6, respectively). Genomic DNA derived from stably transfected and non-transfected (wild-type) control B. bigemina parasites were amplified by PCR using the set primers: A) egfp-Fwd and ef1α-Rev; and B) egfp-Fwd and egfp-Rev ( Fig. 5 and Table 1). The expected band of 2.2 kb was observed only upon amplification of the transfected parasite line bigemina-big-ef-egfp-bsd but not on the gDNA derived from the wild-type B. bigemina parasites (Fig. 5a). The sequence of the 2.2 kb PCR amplicon was consistent with integration of the transfected egfp-bsd gene and its flanking regions into the B ef-1α gene of B. bigemina by homologous recombination (GenBank accession nr: MG234552). In addition, control PCR reactions using primers representing sequences present only in the transfection plasmid (Fig. 5b) (egfp-Fwd and egfp-Rev) only amplify a similar fragment in the transfected line bigemina-big-ef-egfp-bsd and transfection plasmid pbig-ef-egfp-bsd, but not on the gDNA derived from the wild-type B. bigemina parasites.
PCR amplifications performed using similar sets of primers on transfected bovis-big-ef-egfp-bsd and wild-type B. bovis parasites are shown in Fig. 6. Integration PCR using the set of primers egfp-Fwd and Bbov-UpS-efB-Rev yielded a ~2.2 kb band ( Fig. 6a and Table 1) Fig. 6b. Figure 6c also show alignments among the regions of insertion of the transfected genes into B. bovis and B. bigemina transfected parasites. PCR amplifications using primers rap-1a-Fwd and rap-1a-Rev confirmed the presence of the rap-1a gene in gDNA from transfected and wild-type parasites (Fig. 6b). An amplicon of ~320 bp was obtained (Fig. 6b). We also performed immunoblotting to confirm expression of eGFP-BSD by the transfected B. bigemina and B. bovis parasites. As can be seen in the immunoblot, no reactivity was observed for any of the antibodies used and uninfected bovine RBC (Fig. 7a and lanes 5). The wild-type B. bigemina and bigemina-big-ef-egfp-bsd reacted against B. bigemina rap-1 protein (~50 kDa) and the B. bovis and bovis-big-ef-egfp-bsd reacted against B. bovis rap-1 protein (~50 kDa) (Fig. 7b,c) using anti-RAP-1 MAbs as positive controls. Also, anti-GFP antibodies reacted with proteins present in both bigemina-big-ef-egfp-bsd and bovis-big-ef-egfp-bsd parasite lines with the expected molecular weight of the GFP-BSD fusion protein (~40 kDa), but not in wild-type B. bigemina and B. bovis (Fig. 7d).
Taken together, Southern blot, PCR and immunoblot data confirmed stable integration of the egfp-bsd gene into the genomes of both B. bovis and B. bigemina with demonstrated expression of the transfected exogenous genes.

Discussion
Here, a stable transfection of B. bigemina and B. bovis using identical B. bigemina insertion and gene regulatory sequences that can be used for functional gene characterization or for the delivery of exogenous antigens by Babesia spp. parasites is described.
Importantly, the transfected egfp-bsd gene integrated as a single copy in the expected ef-1α locus in transfected B. bigemina parasites and no episomal forms of exogenous DNA bigemina-big-ef-gfp-bsd were detectable in parasites that were selected with blasticidin for at least two months.
A similar pattern of specific integration by means of homologous recombination for the exogenous transfected egfp-bsd gene into the expected ef-1α locus was also found to occur in B. bovis parasites using a plasmid transfection vector designed for integration into the B. bigemina ef-1a locus, despite the occurrence of sequence divergence in the ef-1α gene among the two species. Sequence comparisons among the ef-1α orf of B. bovis and B. bigemina show a level of identity of 87.45%. (Suppl. Figure 2 and Suppl. Table 1). Remarkably, that level of identity was sufficient to allow specific integration of the pbig-ef-egfp-bsd gene into the ef-1α locus of B. bovis genome, generating a hybrid ef-1α molecule in transfected B. bovis. The DNA sequence comparisons between ef-1α locus of B. bigemina and B. bovis (Suppl. Figure 2) provide hints on the mechanisms of homologous recombination operating in the parasite. Additionally, and consistent with previous findings 10 , the ef-1α promoter of B. bigemina was able to generate expression levels of the egfp-bsd gene that are sufficient to sustain growth of transfected parasites at high levels of blasticidin. The strategy for exogenous gene insertion used in this study makes expression of the transfected egfp-bsd gene by the "native" B. bigemina promoter theoretically possible, but this possibility is highly unlikely since the homologous or heterologous ef-1α promoter region (~700 bp), located immediately downstream the truncated ef-1α orf contain numerous stop codons. In addition, we previously demonstrated heterologous promoter function using transient transfection, where the transfected gene is expressed under the control of the sequences present in the transfection plasmid, and without the intervention of the original promoters, supporting the contention that the exogenous integrated promoter is indeed responsible for the expression of the egfp-bsd gene in the stably transfected parasites. Transfected B. bigemina parasites grew three times faster than non-transfected (wild-type) parasites. This might be due to changes in the regulation of the expression of ef-1α locus in the transfected parasites. It will be interesting to determine whether these changes affect the ability and efficiency of the parasite to infect bovine and tick hosts, and whether they are associated with parasite virulence. The ability to transfect B. bigemina and B. bovis using B. bigemina insertion and promoter sequences also has important implications for improving the design of Babesia sp.-based vectored vaccines 6 . On one hand, a vaccine delivery platform based on B. bigemina transfected parasites might be more advantageous compared with B. bovis since the former parasite is known to cause less severe clinical disease in cattle, and, in contrast to B. bovis, does not result in microvascular sequestration of iRBC in the host. In contrast to B. bovis, B. bigemina parasites can be cleared from persistently infected animals, and dual B. bovis-B. bigemina infections are frequent in cattle in endemic areas [15][16][17] .
Possible future applications of this transfection platform include the use of attenuated B. bigemina-transfected parasites that express B. bovis and/or vector tick antigens that induce parasite and/or vector controlling immunity during subclinical persistent infection. These vectored vaccines might become ideal to generate protective immunity effective against both bovine parasites and their vector.
It will also be necessary to test the functionality of the B. bigemina ef-1α promoter in distinct life stages using transfected B. bigemina and B. bovis and parasites. These observations could have practical consequences for developing improved methods for the study and control of these parasites at different life-cycle stages.
In summary, we describe an efficient method for the stable transfection of B. bigemina that can be used for the future development of novel vaccines. These will require assuring that the transfected parasites are safe to deploy and that the genetic modifications do not result in undesirable phenotypic characteristics in potential vaccine candidate strains. The observation that the B. bigemina-specific construct can also be effectively and specifically used