P. falciparum falcipain-2 has received a great amount of attention as a target for therapeutic interventions against malaria, due to its relevant functional role . This work identifies and characterizes bovipain-2, the B. bovis ortholog of falcipain-2. The biological significance of this protein is underscored by the observation that B. bovis growth can be inhibited using cysteine-proteinase inhibitors .
Based on their sequences, falcipain-2 and bovipain-2 are classifiable as cysteine peptidases belonging to Clan CA, subfamily C1A. This peptidase subfamily is characterized by the presence of four catalytic Q, C, H, and A residues present in three separate, well conserved regions of the primary sequence that corresponds to the mature protease, which are known as the eukaryotic thiol (cysteine) proteases cysteine, histidine, and asparagine active site regions (Figure 2). In the final tertiary conformation of the protein, the catalytic amino acids are brought together and constitute the enzyme active site. Importantly, predicted models of the putative catalytic site in falcipain-2 and bovipain-2 showed a similar arrangement of the relevant C, H, and N regions that form the active site of the enzyme in both proteins, despite high sequence polymorphisms in the intervening regions (data not shown). Furthermore, alignment of bovipain-2 orthologs from different apicomplexan parasites showed strict conservation of the catalytic residues and low polymorphism in their surrounding areas (Figure 2). Additionally, full sequence conservation was observed among bovipain-2 protein sequences of B. bovis geographical isolates in the whole region that harbors the active site. This constraint for genetic variation is likely due to the need for keeping the structural conformation of the protein in order to preserve its activity. Thus, taken together these data support a cysteine protease function for bovipain-2, similar to what was previously described for falcipain-2.
Interestingly, apart from bovipain-2, only 3 other C1 cysteine proteinase encoding genes seem to be present in the B. bovis-genome. The bovipain-2 encoding gene corresponds to gene families in T. annulata and T. parva of 7 and 6 members, respectively. In contrast, the other cysteine proteinase encoding genes of B. bovis have a single ortholog equivalent in T. parva and T. annulata. We hypothesize that the expanded cysteine-protease family of bovipain-2 type may have a specific function in T. parva and T. annulata associated with the additional schizont parasite stage and more complex life cycle of Theileria sp. parasites. However, this notion would need to be demonstrated in future investigations. Gene duplications and tandem arrays of similar isoforms of clan CA peptidases have also been found in other related protozoan parasites, but the biological implications of this phenomenon remain unclear .
In P. falciparum falcipain-2 localizes in the food vacuole, where hemoglobin digestion takes place . However, falcipain-2 expression was detected also outside of the food vacuole and near the erythrocyte membrane skeleton which is consistent with the proposed involvement of erythrocyte membrane skeletal protein cleavages in merozoite egress . The punctate immunofluorescence pattern shown in Figure 6 indicates that, similar to its P. falciparum ortholog, bovipain-2 also appears to be localized in a yet undefined, internal organelle of the parasite. Food vacuoles have been described in Babesia sp. but only on the basis of morphological evidence [32, 33]. Thus, further investigations need to be carried out to more exactly elucidate the intracellular location of bovipain-2. However, similarly to Plasmodium, anti-bovipain-2 antibodies also reacted with the cytoplasm of B. bovis-infected erythrocytes of different strains tested, while smears of non-infected erythrocytes showed no reactivity. These observations suggest first, that bovipain-2 and erythrocyte proteases do not share cross-reactive B-cell epitopes and also, that bovipain-2 is released into the erythrocyte cytoplasm. Occasionally, some erythrocytes in which no parasites could be detected showed cytoplasmic reactivity in smears of B. bovis-infected erythrocytes (data not shown). Since non-infected erythrocyte smears showed no reactivity with anti-bovipain-2 antibodies, a possible explanation for this observation is that the parasites egressed from the reactive erythrocytes before the smears were prepared, leaving released proteins behind. Reactivity towards empty erythrocytes of antibodies that recognize babesial proteins was previously observed in the case of the merozoite spherical body protein, Bb-1, which can be found on the cytoplasmic side of the erythrocyte membrane . It has been suggested that Bb-1 is secreted by B. bovis into the erythrocyte cytoplasm, where it might be involved in invasion and/or exit processes.
Anti-bovipain-2 antibodies cross-reacted with B. bigemina merozoites. A 48.5% identity in the predicted amino acid sequences of bovipain-2 and a putative B. bigemina protein named babesipain [35, 36] (Table 2) could be responsible for the observed cross-reactivity. The cytoplasm of B. bigemina-infected erythrocytes also reacted with anti-bovipain-2 antibodies, suggesting that release of babesipain from B. bigemina merozoites has also taken place.
Expression of cysteine proteases in cultured, wild type or attenuated B. bovis parasites can be affected by distinct possible selective pressures. This is significant since the pattern and level of expression of cysteine proteases were also regarded as potential factors affecting the virulence of parasites . Sera from bovines experimentally infected either with the BboS2P or BboR1A Argentinean strains reacted with the recombinant bovipain-2, confirming in vivo expression and immunogenicity of the protein in the blood stages of the parasites regardless of their degree of virulence.
However, because four out of six sera from infected animals tested failed to recognize recombinant bovipain-2, the results also suggest that bovipain-2 might be poorly immunogenic during acute infection. It has been reported that subdominant antigens may prove to be more effective as vaccine candidates than immunodominant antigens . Based on this supposition, bovipain-2 constitutes an interesting candidate for subunit vaccine development. Alternatively, it is possible that these sera also contain antibodies reactive with conformational epitopes in the native protein that are not able to recognize the recombinant version of bovipain-2.
In silico analysis of the B. bovis predicted proteome using the MEROPS database has shown the presence of 66 proteases, that belong to the cysteine (n = 18), serine (n = 18), metallo (n = 19), threonine (n = 6) and aspartic (n = 5) classes (manuscript in preparation). Although this number is small compared to P. falciparum (n = 93), the high number of protease-encoding genes appears to indicate that these enzymes participate in critical metabolic processes and/or signalling mechanisms that need to be separately regulated. On-going experiments to analyze which of these proteases are transcribed in the merozoite and other stages of the parasite life cycle will throw light on their possible functional roles and their suitability as targets for improved control methods for bovine babesiosis.