The B. bovis, B. bigemina, and B. ovata RRA sequences (XP_001610950.1, XP_012766682.1, XP_028866572.1, respectively) were modeled with AlphaFold Data (2022) DeepMind Technologies Limited (23, 24). In addition, molecular graphics and analyses were performed with UCSF ChimeraX, developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (25).

Database sequence searches homologies were performed using BlastP (Protein BLAST: search protein databases using a protein query (nih.gov)). Multiple sequence alignments were performed using Clustal Omega (Clustal Omega < Multiple Sequence Alignment < EMBL-EBI)

Partial size rra genes from DNA extracted the B. bovis Mo7 (26), S79-T3BO (27), Argentina (28) RRA and Australia strains (29) were amplified using primers: RRA-F (5’ GGGCAATGTGCTATCTTGGATACTG 3’) and RRA-Rev (5’ CAAACTCTGTCGTAAAGGTGCCCG 3’):. All amplicons were cloned into pCR-TOPO 2.1 vector (Invitrogen, CA, USA) and sequenced in full by Eurofins MWG Operon (Louisville, KY).

The Mo7 biological clone of B. bovis (20) was used as a source of antigen for immunoblots and immunofluorescence analysis. Parasites were grown in long term microaerophilous stationary-phase culture by previously described methods (20).

Pre-immune and immune sera from the B. bovis experimentally infected cattle used in the ELISA analysis were generated in a previous study (30, 31).

RRA synthetic peptides were obtained from Pacific Immunology Corp. (Ramona, CA). The sequence of the peptides used in this study is as follows: RRA NT: PTIRLPNTYQLEAAF (>80% purity as determined by HPLC analysis); RRA CT: PVKWAVNLIENDDAPYGW (>80% purity as determined by HPLC analysis); The B. bovis derived CT-peptide representing a B-cell epitope of RAP-1 (12) was produced by Vivitide (Louisville, KY, USA): PTKEFFREAPQATKHFLDENIGA (>95 purity as determined by HPLC analysis). Rabbit antibodies were generated against the peptide 15-mer RRA-NT B. bovis: PTIRLPNTYQLEAAF; The peptide 15-mer RRA NT was conjugated to keyhole limpet hemocyanin (KLH) and used in the immunization of rabbits by Pacific Immunology Corp. (Ramona, CA). The specificity of the rabbit antibody against the 15mer peptide was also tested by iELISA using synthetic 15mer and control peptides coated into ELISA plates as described in section 2.4, below. The specificity of the monoclonal antibodies reactive against the B. bovis RAP-1 CT Babb75 was previously described (12).

The full size B. bovis RRA gene (XP_001610950.1) and RAP-1NT (12, 14) sequences were codon optimized for expression in prokaryotic cells, synthesized, and cloned into vector pET-30a(+) with 6His tag for protein expression in E. coli BL21 star (DE3) cells. After expression, the proteins were obtained from inclusion bodies and purified by Ni-column with a purification rate of >85%. The starting volume was 1 L of Terrific medium broth, with a final yield of 18.13 mg. Level of protein purity and expected molecular weight were determined by SDS–PAGE and Western blot. Briefly, the proteins were separated in 4–20% Precast TGX gel (BioRad, Hercules, CA, USA) and stained with Coomassie blue using a standard protocol ( Supplementary Figure 1 ). For Western blot analysis, the SDS–PAGE separated proteins were transferred to a PVDF membrane using manual transfer and probed with HRP-conjugated His-Tag Monoclonal antibody (Proteintech, Rosemont, IL, USA). The size marker used in SDS–PAGE and Western blot was Prestained Precision Plus Protein Dual Color Standard (BioRad, Hercules, CA, USA). Identity confirmation of rRRA protein was performed by LC–MS/MS, as per the third-party company’s protocol (GenScript, Piscataway, NJ, USA). The recombinant protein was stored at −80 °C in a 50 mM Tris-HCl, 150 mM NaCl, 10% Glycerol, 0.2% SDS, pH 8.0 buffer until further use.

Indirect ELISA (iELISA) for synthetic peptides were performed on 96-well Immulon™ 2HB microtiter plates (Thermo Fisher Scientific, Waltham, MA) coated overnight at 4°C with 50 μl of synthetic peptides RAP-1 CT, RAP-1 NT, RRA NT, RRA CT, and randomized RRA-NT peptide (sequence: DQTAINREPYPTALL) (2 μg/ml). Plates were then washed three times using 200 μl blocking buffer (0.2% I-Block™ in 1xPBS with 0.1% Tween 20) and blocked with 300 μl of the same buffer for one hour at room temperature. After blocking, primary antibody was diluted 1:10 for IgG, or 1:20 for IgM detection, and 50 μl was added to have technical duplicate wells. For specificity of the rabbit antibody against the RRA NT 15mer peptide, the primary immune and pre-immune rabbit antibodies were diluted 1:500 in blocking buffer. Plates were incubated for one hour at room T° and then washed five times in 200 μl blocking buffer. Then, 50 μl of a 1:500 dilution of anti-bovine total IgG or 1:300 dilution of anti-bovine total IgM peroxidase labeled secondary antibodies (SeraCare, Milford, MA) were added to each well, and plates were incubated for 45 minutes at room T°. For the plates incubated with rabbit primary antibodies, the secondary goat labeled secondary antibodies anti-rabbit IgG was diluted 1:1000 in identical buffer. After incubation, plates were washed four times using 200 μl ELISA washing buffer and two times with 200 μl PBST buffer. Following, 55 μl of SureBlue™ TMB (SeraCare, Milford, MA) was added to each well and plates were incubated for 10 minutes. After addition of 55 μl TMB stop solution (SeraCare, Milford, MA), absorbance was measured at 450 nm using the SpectraMax^® 190 plate reader (Molecular Devices, San Jose, CA). The serum samples analyzed hereby were generated in a previous study (30, 31) involving vaccination of cattle with the in vitro culture attenuated strain of B. bovis Att- S74-T3Bo B. bovis that was challenged with a virulent strain of the parasite Vir-S74-T3Bo B. bovis. Statistical significance was determined using a paired 2 tails type 1, T-test. The previous animal study was reviewed and approved by Washington State University Institutional Animal Care and Use Committee (IACUC protocol number 2020-51).

Briefly, total antigens (2 μg/lane) from uninfected bovine RBC or Mo7 iRBC (2% percentage of parasitized erythrocytes, PPE) were ran through 4-20% TGX™ gels (Bio-Rad Laboratories, Hercules, CA), using 5X Sample Buffer containing 10% 2-Mercaptoethanol (GenScript, Piscataway, NJ, USA). Antigens were then transferred to nitrocellulose membranes using iBlot 2™ (Invitrogen, Waltham, MA), and the membranes were blocked 1 hour in 5% milk. Membranes were then washed one time in 1xPBS with 0.1% Tween 20 (PBS-T) and incubated with a 1:500 dilution of either pre-immune rabbit sera or rabbit anti 15-mer antibody serum in 5% milk for one hour rocking at room temperature. After that, membranes were washed three times using PBS-T and incubated with a 1:1000 dilution of anti-rabbit IgG peroxidase labeled secondary antibodies (SeraCare, Milford, MA) in 5% milk for 30 minutes rocking at room T°. Membranes were then washed again three times in PBS-T and incubated with Prometheus Protein Biology Products 20-300B ProSignal^® Pico ECL Reagent (company)?. Visualization of immune complexes was performed using the Azure™ Imaging System (Azure Biosystems, Dublin, CA). The immunoblot in Figure 1C was performed using 1:500 anti-His tag antibody (Bio-Rad Laboratories, Hercules, CA), followed by peroxidase labeled anti-Mouse antibody 1:5000 (SeraCare, Milford, MA).

Extraerythrocytic free merozoites were isolated from cultured B. bovis Mo7 parasites using centrifugation, as previously described (32). The samples were then washed in 3% bovine serum albumin (BSA) and aliquoted for use in cell permeabilized and non-permeabilized indirect immunofluorescence assays (IFA). For permeabilized IFA, samples were first smeared on a slide, fixed for 5 min in 100% acetone, and then incubated with 0.1% Triton X-100. The slides were then incubated in 10% BSA for 1 h with a combination of anti-RAP-1 antibodies (2 ug/mL) and anti-RRA peptide antibodies (1/500) (31). The slides were washed three times with PBS and incubated in 10% BSA with goat anti-rabbit IgG Alexa Fluor^® 555 (red fluorescence) and goat anti-mouse IgG Alexa Fluor^® 488 (green fluorescence) (Thermo Fisher Scientific, Waltham, MA). The slides were then washed three times with PBS and mounted with a drop of Prolong™ Gold Anti-fade with 4′,6-diamidino-2-phenylindole (DAPI) (Thermo Fisher, Waltham, MA, USA) and cover slip. Non-permeabilized samples were incubated with the antibodies, washed within a 1.5 mL tube, then fixed on a slide, and mounted in an identical manner as the permeabilized samples. The slides were analyzed using a Leica SP8-X White Light Laser point scanning confocal microscope (Leica Microsystems, Wetzlar, Germany). The digital images were processed using Leica LAS X analysis software (Leica Microsystems, Wetzlar, Germany) to produce individual and merged images.

Sequence comparisons among the RRA proteins from B. bovis, B. bigemina, and B. ovata, revealed strong conservation of the 15-mer sequence that overlaps and extends in one residue the 14-mer motif previously identified as conserved among canonical RAP-1 proteins (11, 13) ( Figures 2A, B ).

A detailed comparison of the 15-mer conserved motif sequences among the RRAs from B. bovis, B. bigemina, B. ovata, and B. bovis RAP-1, is shown in Figure 2B . At least 13 out of the 15 amino acids in this sequence are strictly conserved among the RRA sequences from these three distinct selected Babesia species, and 8 residues are also strictly conserved in B. bovis RAP-1 ( Figures 2B, C ). Furthermore, the two variable residues in these sequence comparisons (HxR; DxN) are conservative substitutions (33). In addition, and in all cases, the identities among the RRA 15-mer motifs among these three parasites are significantly higher than the identities among their RRA full size amino acid sequences ( Figure 2C ), suggesting that the region including the conserved motif has a lower evolutionary rate than other regions in the RRA gene. However, RRA, including the 15-mer motif, is fully conserved among B. bovis strains from Argentina, Mexico, Australia, as well as in the clonal Mo7 strain ( Supplementary Figure 1 ). Also, as previously observed (26), the 15-mer motif is also well conserved among RRA and RAP-1 of B. bovis ( Supplementary Figure 2 ).

Interestingly, BlastP database searches using just the conserved 15 amino acid motif region as a query revealed strong similarity with sequences present in at least two distinct adenylate cyclase type 10-like proteins (GenBank: KMQ94175 and XP_029155514.1) (red boxes, Supplementary Figures 3 , 4 ), as well as partially conserved in other ATP-binding proteins, such as dynein (ie: GenBank XP_019924815), an ATP-binding cytoskeletal motor protein. An alignment of the full B. bovis RRA amino acid sequence with the sequence of the adenylate cyclase containing homology to the 15-mer motif (AC1) using CLUSTALW revealed overall 22.7% of identity among RRA and AC1 ( Supplementary Figures 3 , 4 ). However, the highest similarity among these two proteins is present in the region representing the 15-mer motif (overall 67% identity, with 2 conservative or semiconservative substitutions in a 13 amino acid stretch, as shown in Supplementary Figure 3 ). Furthermore, the sequence homology extends along both NT and CT ends of RRA and, significantly, includes also one of the highly conserved cysteine residues ( Supplementary Figures 3 , 4 ). Consistently, BlastP database searches using just the 13 amino acid regions from AC that almost fully matches the 15-mer conserved RRA sequence as a query ( Supplementary Figure 3 ), produced significant hits with the RRA of B. bovis and B. bigemina, and other ATP-Binding (ABC) proteins (data not shown).

In addition, comparisons of AlphaFold predicted surface structures of RRA from B. bovis, B. bigemina, and B. ovata ( Figure 3A ), suggest overall structural conservation among RRA proteins despite their amino acid sequence differences ( Figure 2A ). Interestingly, the region including the 15-mer motif in RRA of B. bovis, B. bigemina and B. ovata, have identical predicted structures, consisting of α-helix and β-sheet regions joined by a non-structured linker ( Figure 3B ). The 15-mer motif of RRA described hereby is a one residue extension version of the 14-mer motif in RAP-1 proteins (34) which contains its more conserved residues buried in the core of the protein, which overall has a globin-like structure. Remarkably, this includes the key L-122 and P-123 residues of the 15-mer motif that are conserved on other pRAP-1 superfamily structural homologues identified in more distant species such as Plasmodium and Toxoplasma (34). However, residues in the relative positions 1, 2, 4, and 7 (differentially colored in Figure 3A and highlighted in Figure 2B ) of the 15-mer conserved region are predicted to be exposed on the surface of the RRAs derived from the 3 parasites compared hereby. Thus, according to these structural AlphaFold models, at least some residues of the conserved 15-mer of the RRA molecules might be exposed on the surface of the molecules ( Figure 3B ). According to this prediction, there are four exposed residues in matching positions for the RRAs of the three parasites ( Figure 2A and highlighted residues in Figure 2B ), except for B. ovata, which has an additional predicted surface exposed residue (T-113). While the matching 15-mer surface exposed predicted residues P and T are fully conserved among the three molecules ( Figures 2A, B ), the other two matching residues (Positions 4 and 7 in Figure 2B ) are the only ones which are variable among the three RRA molecules compared in this study. Interestingly, two out of these four predicted surface exposed amino acid residues (#4 and #7 in Figure 2B ) are the only ones that vary in the 15-mer motif sequences among these three RRA molecules. Notably, as pointed out before, the amino acid changes at these two positions can be considered conservative substitutions (33) which do not likely alter the structural features of this region and may have no significant influence on the functional role of the motif. In contrast to the NT regions, and consistent with previous findings (34), the CT end regions of these three RRA proteins do not show AlphaFold-predicted species-conserved structures at all ( Figures 3A, B ). Interestingly, AlphaFold surface structure predictions for the Adenyl-Cyclase and Dynein molecules which also contain versions of the 15-mer, as described above, also show globular shapes which appear to be similar to the RRAs, with residues of their 15-mer-like sequences also exposed on their surfaces ( Figure 3A ).

Altogether, the data implies that the conserved 15-mer motif may be critical for RRA, as well as for RAP-1, functions and suggests that it may have surface exposed residues that might be the target of invasion-neutralizing antibodies.

To analyze immunogenicity and confirm exposure conserved motif, we first generated rabbit polyclonal antibodies against a peptide representing the 15-mer motif of the B. bovis RRA. These antibodies were used in ELISA, immunoblots and immunofluorescence analysis of infected erythrocytes and extraerythrocytic merozoites using permeabilized and non-permeabilized B. bovis parasites. In addition, we also examined whether antibodies in B. bovis protected calves recognized B-cell epitopes that might be in the B. bovis 15-mer conserved motif.

The sera from a rabbit immunized with the 15-mer motif of RRA, but not pre-immune rabbit sera, strongly reacted with a synthetic peptide representing the 15-mer motif of RRA, but not with a randomized version of the same peptide in an ELISA test ( Supplementary Figure 5A ), confirming its specificity. In addition, the rabbit antibodies against the 15-mer of RRA also reacted with recombinant RRA antigen (rRRA) generating a ~38 kDa band which matches the predicted size of this recombinant protein, but do not react with rRAP-1N, a recombinant antigen that represents the conserved region of B. bovis RAP-1 that includes the 15-mer sequence described in Figure 2B ( Figure 1 , Supplementary Figure 5B ). However, as shown in the Figures, the rabbit antibodies against the 15-mer of RRA also recognized a HIS-containing ~78 kD recombinant protein that was co-purified with rRRA. This pattern of reactivity was similar to the one obtained using anti-HIS antibodies as the primary antibody in immunoblots ( Figure 1C , Supplementary Figure 5C ), which include reactivity with a ~78 kDa protein, suggesting that the larger recombinant antigen may represent a RRA dimer. The rabbit antibodies against the 15-mer motif of RRA only recognized a native ~78 kDa B. bovis protein, in a lysate of B. bovis infected erythrocytes derived from in vitro cultured parasites in immunoblot analysis ( Figure 1B ). Consistent with what was found for rRRA, this pattern of reactivity suggests that the native ~78 kDa protein recognized by the rabbit anti15-mer RRA antibodies may represent a dimer of native RRA. Alternatively, it is possible that the amounts of monomeric RRA, but not of the dimeric form, present in the culture lysates, is below the limit of detection of the immunoblot method as it was suggested in previously published studies (22). It is also possible that the larger ~80 kDa band represents post-translationally modified native RRA, or a complex of RRA with an unknown molecule. No reactivities were detected for the rabbit anti-15-mer antibodies against proteins in a non-infected bovine erythrocyte lysate, and no antigens in the lysates of infected and non-infected erythrocytes were recognized by rabbit pre-immune sera ( Figures 1A, B ).

In addition, we compared the pattern of immunofluorescence reactivity of the anti-15-mer motif rabbit antibody (green fluorescence) with the monoclonal antibody BABB75, reactive with B. bovis RAP-1 (red fluorescence) (12). The immunofluorescence results demonstrated reactivity of the rabbit anti-15-mer motif antibody with an antigen expressed in permeabilized infected erythrocytes (green fluorescence) ( Figure 4A ) and with the surface of non-permeabilized B. bovis extra-cellular merozoites (green fluorescence) ( Figure 4B ). These immunofluorescence images also suggest co-localization of both antigens (orange/yellowish fluorescence) in most intracellular merozoites. A more detailed view of the pattern of reactivity of the monoclonal anti-RAP-1 antibody BABB75 (red fluorescence), and anti 15-mer RRA peptide (green fluorescence) rabbit antibodies with permeabilized infected erythrocytes is shown in panels 11-15 of Figure 4A , depicting a merozoite invading an erythrocyte. The RAP-1 protein, recognized by the mAb BABB75 (red fluorescence), is clearly expressed in the apical end of the parasite during erythrocyte invasion, however the reactivity of anti-RAP-1 and the anti-15-mer antibodies also overlap in the apical end (orange/yellowish fluorescence), also confirming the co-expression of both molecules and a possible role of the neutralization-sensitive RRA (22), during erythrocyte invasion.

The analysis performed on non-permeabilized extracellular merozoites is shown in Figure 4B where anti-RAP-1 and anti-RRA 15-mer antibodies reacted with the surface of aggregated extracellular B. bovis merozoites. Overall, the data supports the expression of merozoite surface exposed epitopes that are recognized by the antibodies against the 15-mer motif peptide, leading to the notion that the predicted exposed amino acids of the 15-mer motif could be a part of a merozoite surface exposed B-cell epitope.

We then examined whether B-cell epitopes in the conserved 15-mer region of RRA are also recognized by antibodies in cattle protected from B. bovis infection. To this end, we analyzed IgM ad IgG responses in sera derived from calves and adult cattle that were vaccinated with in vitro culture attenuated parasites, and then challenged with a virulent strain of B. bovis in previous studies (30, 31). We compared antibody responses against a synthetic peptide representing the sequence of the 15-mer motif of the B. bovis RRA, and a peptide representing an immunodominant B-cell epitope expressed in the CT end of RAP-1, recognized by monoclonal antibody BABB75 (12). Antibodies in the calf sera were examined at different dates following immunization with an attenuated live vaccine and challenge with virulent B. bovis parasites (30, 31).

The IgM responses are represented in Figure 5 . While both adult and young cattle mounted responses against the RAP-1CT and the RRA 15-mer peptide, the IgM responses are much higher for the 3 calves analyzed hereby (p< 0.05). ( Figure 5A ). These calves also mounted IgM responses recognizing the RRA 15-mer peptide, specially, calf C1735 (blue line, Figure 5B ) but at much lower levels compared to the response against the immunodominant RAP-1CT peptide (p < 0.05). Regarding the IgM responses of adult cattle, although both groups generated IgM responses, there is an interesting clear and anamnestic response against the RRA NT 15-mer peptide by cow C1707 (Purple line Figure 5D ).

The IgG responses in calves and adult cattle are represented in Figure 6 . Both calves and adult cows mounted strong IgG responses against the RAP-1CT peptides, but the responses against this peptide are more consistent and uniform among calves. All three analyzed calves showed strong IgG responses against the RAP-CT peptide, and at least two calves, C1735, and C1738 (blue and brown lines, respectively, in Figure 6B ), showed significant, albeit lower, IgG responses to the 15-mer RRA NT peptide (p <0.05). Two adult cows (C1703 and C1705 represented with grey and orange lines respectively, in Figure 6C ) mounted strong IgG responses against the RAP-1CT peptide. Interestingly, and similar to what was found for the IgM responses, there was a strong anamnestic response for cow C1707 against the peptide representing the conserved RRA 15-mer peptide ( Figure 6D ).

Altogether, the iELISA data demonstrated the presence of at least one B-cell epitope that is recognized by antibodies in B. bovis infected adult and young cattle. Regardless of the quantitative differences found, we concluded from these experiments that at least one B-cell epitope is present in the 15-mer region of RRA. However, it must be noted that the magnitude of the humoral immune responses against Rap-1 CT, a known immunodominant epitope, are much higher than the responses against the epitope in the RRA 15-mer, indicating that the B-cell epitope in the RRA conserved 15-mer region is likely sub-immunodominant.