Occurrence and Genetic Diversity of Babesia caballi and Theileria equi in Chilean Thoroughbred Racing Horses

This study aimed to serologically and molecularly survey Babesia caballi and Theileria equi in thoroughbred horses from racecourses in Chile. Additionally, the genetic diversity of the positive samples was assessed. A total of 286 thoroughbred horses from the Santiago and Valparaíso racecourses had their serum samples submitted to an ELISA for B. caballi and T. equi, and 457 samples (from the Santiago, Valparaíso, and Concepción racecourses) were tested with nested PCRs for the B. caballi 48 KDa rhoptry protein (RAP-1) and T. equi 18S rRNA genes. Selected RAP-1 and 18S positive products were sequenced to perform phylogenetic and haplotype analyses. An overall seroprevalence of 35.6% was observed for these Chilean racecourses: 23.7% for T. equi, 8.4% for B. caballi, and 3.5% for both agents. Overall, a 53.6% occurrence by nPCR was detected for the three Chilean racecourses: 44.2% for T. equi, 5.4% for B. caballi, and 3.9% for both agents. Phylogenetic analysis of T. equi and B. caballi showed genetic proximity with sequences previously detected in other countries. Haplotype analysis revealed a low diversity among the Chilean sequences, which may have originated from those reported in Brazil, Israel, or Cuba. Babesia caballi and T. equi were detected for the first time in Chilean thoroughbred horses.


Introduction
Equine babesiosis and theileriosis, together known as equine piroplasmosis (EP), are two tick-borne diseases caused by the protozoans Babesia caballi and Theileria equi (Apicomplexa: Piroplasmida) [1], which belong to the Babesiidae and Theileriidae families, respectively [2]. EP has high morbidity and mortality rates and is, therefore, a serious concern in the equine industry [3]. Babesia caballi and T. equi are transmitted by ticks (biological vectors), and frequently co-infect horses [4][5][6][7]. Although transmission via the placenta from pregnant mares to fetuses has been described, this route is of little epidemiological significance for EP, and is considered rare, especially for B. caballi [8][9][10][11]. Additionally, B. caballi and T. equi can be transmitted mechanically or iatrogenically via contaminated needles and syringes, blood transfusions, and surgical instruments [9].
Both B. caballi and T. equi infections can be subclinical to severe acute diseases, and symptoms are often variable and non-specific in horses [5,12]. In general, infection with T. equi causes a more severe clinical disease than infection with B. caballi [13]. Although

Phylogenetic and Haplotype Analyses
Out of the 202 positive samples for T. equi that were tested by targeting the complete 18S rRNA gene (1600 bp) [37], 10 amplicons that had good band quality in agarose gel were selected and submitted for sequencing (gene bank accession numbers: MT463604-MT463613) and maximum likelihood analyses (ML). These sequences, when analyzed by BLASTn (using default parameters), showed a shared identity ranging from 99% to 100% with T. equi sequences previously detected from equids in Israel (MN611352.1), Brazil (MG052898.1), Turkey (MG569905.1), and the Philippines (KX900447.1). Among the 25 positive samples for B. caballi, only one sample (MT521699) could be sequenced using the 48 KDa rhoptry protein gene protocol [38], and submitted for Bayesian analyses. The BLASTn analyses (using default parameters) revealed that this sequence shared 99% to 100% identity with B. caballi amplified from equine blood samples from the Philippines (KX900447.1), Egypt (KR811095.1), Mongolia (AB703053.1), and Cuba (KY111765.2). All sequences amplified in the present study showed query coverage ranging from 99% to 100%, and an E-value of 0.
While the maximum likelihood analysis of T. equi showed four clusters in the phylogenetic tree, the T. equi sequences (n = 10) detected in this study were all positioned in cluster #3 (Figure 1). The Chilean sequences formed a cluster with each other and with sequences found in Israel, Brazil, and Cuba ( Figure 1). Bayesian analysis of B. caballi formed six clusters, and the sequence detected in this study was closely positioned to sequences previously detected in horses from Egypt and Mongolia (cluster #2) ( Figure 2).  Out of 35 T. equi 18S rRNA sequences used for the haplotype analysis (10 from horses in this study and 25 from horses worldwide), 17 different haplotypes were detected, with a haplotype diversity of 0.934 ± 0.021 and a nucleotide diversity of 0.02186 ± 0.00196 (Table   Out of 35 T. equi 18S rRNA sequences used for the haplotype analysis (10 from horses in this study and 25 from horses worldwide), 17 different haplotypes were detected, with a haplotype diversity of 0.934 ± 0.021 and a nucleotide diversity of 0.02186 ± 0.00196 (Table Out of 35 T. equi 18S rRNA sequences used for the haplotype analysis (10 from horses in this study and 25 from horses worldwide), 17 different haplotypes were detected, with a haplotype diversity of 0.934 ± 0.021 and a nucleotide diversity of 0.02186 ± 0.00196 (Table 2). Four haplotypes were detected in horses from this study, which originated from those previously detected in Brazil and Israel (haplotype #4), separated from them by a few mutational events. Haplotypes #6, #16, and #17 from Chile originated from the same haplotype, also from Chile (#1). Out of the four T. equi haplotypes detected in Chile, two were exclusive to this country, and the other two have also been previously found in Brazil, Cuba, and Israel ( Figure 3). 2). Four haplotypes were detected in horses from this study, which originated from those previously detected in Brazil and Israel (haplotype #4), separated from them by a few mutational events. Haplotypes #6, #16, and #17 from Chile originated from the same haplotype, also from Chile (#1). Out of the four T. equi haplotypes detected in Chile, two were exclusive to this country, and the other two have also been previously found in Brazil, Cuba, and Israel ( Figure 3). On the other hand, 17 B. caballi sequences (one from this study) were used for haplotype analysis, resulting in 14 different haplotypes, showing a haplotype diversity of 0.971 ± 0.032 and a nucleotide diversity of 0.21122 ± 0.082 ( Table 2). The only B. caballi haplotype detected in Chile (#2) has also previously been found in Egypt and Mongolia ( Figure 4). The haplotype from Chile originated from one previously detected in Cuba (haplotype #3), after one mutational event.  On the other hand, 17 B. caballi sequences (one from this study) were used for haplotype analysis, resulting in 14 different haplotypes, showing a haplotype diversity of 0.971 ± 0.032 and a nucleotide diversity of 0.21122 ± 0.082 ( Table 2). The only B. caballi haplotype detected in Chile (#2) has also previously been found in Egypt and Mongolia ( Figure 4). The haplotype from Chile originated from one previously detected in Cuba (haplotype #3), after one mutational event.

Discussion
While, in South America, B. caballi and T. equi DNA has been detected in horses from Brazil [28,29] and Venezuela [24], B. caballi and T. equi have only been molecularly reported in Chile in one study with a small sample of horses (n = 77) from the Metropolitan region [34]. To the best of our knowledge, this is the first molecular survey of B. caballi and T. equi in racehorses from Chile, as well the first assessment of piroplasmid haplotypes in horses from this country.
The overall molecular occurrence for EP agents obtained in thoroughbred horses from Chilean racecourses (B. caballi [ [40,41]. As a global trend [9], T. equi is more prevalent than B. caballi, whereas coinfection is less frequent. The higher prevalence of T. equi when compared to B. caballi is in accordance with the findings of other authors [42][43][44], which can be attributed to the lifelong persistence of T. equi infection, compared with the effective elimination of B. caballi [6,45]. Interestingly, the Club Hípico de Concepción had a higher molecular prevalence of these agents (B. caballi [ [47]. The higher occurrence of equine piroplasmids in the Club Hípico de Concepción might be related to an infection cluster caused by the management of horses (favoring transmission) and their origin (favoring the introduction of asymptomatic carriers), resulting in high numbers of asymptomatic animals [48]. It is recommended that this racecourse investigate its management of horses, the presence of ectoparasites, contaminated fomites, and environmental conditions. The

Discussion
While, in South America, B. caballi and T. equi DNA has been detected in horses from Brazil [28,29] and Venezuela [24], B. caballi and T. equi have only been molecularly reported in Chile in one study with a small sample of horses (n = 77) from the Metropolitan region [34]. To the best of our knowledge, this is the first molecular survey of B. caballi and T. equi in racehorses from Chile, as well the first assessment of piroplasmid haplotypes in horses from this country.
The  [40,41]. As a global trend [9], T. equi is more prevalent than B. caballi, whereas co-infection is less frequent. The higher prevalence of T. equi when compared to B. caballi is in accordance with the findings of other authors [42][43][44], which can be attributed to the lifelong persistence of T. equi infection, compared with the effective elimination of B. caballi [6,45].
Interestingly  [47]. The higher occurrence of equine piroplasmids in the Club Hípico de Concepción might be related to an infection cluster caused by the management of horses (favoring transmission) and their origin (favoring the introduction of asymptomatic carriers), resulting in high numbers of asymptomatic animals [48]. It is recommended that this racecourse investigate its management of horses, the presence of ectoparasites, contaminated fomites, and environmental conditions.
The  [23,[55][56][57]. Moreover, worldwide serological prevalence might underestimate a novel species of Theileria, detected in horses from the USA. The recently described piroplasmid T. haneyi [58] is undetectable using current surveillance tools that target EMA-1 of the immunodominant EMA superfamily, such as the ELISA used in our study. Here, we used EMA-1 ELISA; however, previous serosurveys in thoroughbred horses from Chile have been based on Agar Gel Immunodiffusion Assay (AGID), complement fixation test (CFT), and indirect immunofluorescence antibody test (IFA) [20,32,33,35], which make specific comparisons between prevalence rates difficult.
We We did not observe any relationship between serology and positive PCRs, reflecting the different stages of the EP disease and the characteristics of each methodology. PCR is the best technique for diagnosing animals during acute infections, when there are no detectable levels of antibodies. It is sensitive to the detection of the DNA of the parasite in early stages of the infection, being also able to detect carrier animals [9]. On the other hand, ELISA will detect antibodies, and is therefore a good test for chronic infections [7,9,23,28,55,56].
Preliminary studies of the 18S rRNA gene allowed for the detection of five T. equi groups, named A to E [9,61], of which four were detected in this work [9,37]. All T. equi samples from Chile belonged to a single group based on 18S rRNA gene sequences, and were assigned to Group A from Bhoora et al. (2009) [37]. In contrast, previous studies have shown higher genetic diversity within 18S rRNA sequences from T. equi in horses from Brazil [29], Jordan [62], South Africa [37], and Switzerland. A low diversity within T. equi sequences from Chile could be related to the history of horses' arrival in the country, and the equid business movement trends [63]. Consistent with the above, although Chile has one of the most prestigious racecourses in Latin America [60], it is a minor participant in the global equine industry, and engages in less movement of horses [64], which in turn might have contributed to the low T. equi genetic diversity observed in the present study. Moreover, although Ixodidae ticks are present in Chile [65], we did not find ticks parasitizing the horses sampled in this study.
The B. caballi phylogenetic analysis divided the sequences into four genotypes, in agreement with previous studies [55,66]. Even though the Chilean sequence belonged to the same group as sequences from Cuba, Egypt, and Mongolia, further studies are needed to determine the genetic variations between B. caballi isolates from Chile.
Herein, both T. equi and B. caballi haplotypes presented genetic proximity with sequences previously detected in other countries. Interestingly, the sequences detected in Chile might have originated from those reported in Brazil, Israel, or Cuba. Indeed, they were separated from each other by a few mutational events, which might indicate a close epidemiological link among isolates [67]. These findings also pointed out that the genetic diversity of equine piroplasmids is not geographically delimited [55]. Finally, the haplotype analysis of our sequences with others around the world showed a high level of genetic diversity in B. caballi (Hd: 0.971 ± 0.032) and T. equi (Hd: 0.934 ± 0.02) haplotypes, which have been observed worldwide [7,68,69]. This can be explained by the biphasic modes of piroplasmid reproduction (sexual and asexual), and the intensity of transmission, influenced principally in our country by mechanical or iatrogenic transmission via contam-Pathogens 2021, 10, 714 8 of 15 inated needles and syringes, blood transfusion, and surgical instruments [9]. Furthermore, comparison of 18S rDNA sequences classified as T. equi from multiple locations worldwide indicates greater genetic diversity than expected within a single species. This suggests that multiple taxa are likely to be involved, including additional, so far undescribed and novel species, as was the case for T. haneyi [58].

Animals and Areas of Study
This study was approved by the Universidad Austral de Chile (UACh) bioethics committee under UACh 238/2015. To accurately determine the occurrence of B. caballi and T. equi at the Concepción, Santiago, and Valparaíso racecourses, the required sample size was estimated according to Thrusfield (2007) [70]. The estimated prevalence was 10%, based on a preliminary study in Chile carried out by Vargas et al. (2004), with a precision of 5% and a 95% confidence interval. The minimal sample required was 139 horses from each racecourse. Each owner signed a consent form before the horses' blood samples were taken.
Over 12 months (February 2015 to February 2016), 457 thoroughbred horses had their blood sampled by a veterinary team. One hundred and fifty-one horses came from the "Club Hípico de Santiago" racecourse (Santiago City, Central Chile: 33 •  which have been observed worldwide [7,68,69]. This can be explained by the biphasic modes of piroplasmid reproduction (sexual and asexual), and the intensity of transmission, influenced principally in our country by mechanical or iatrogenic transmission via contaminated needles and syringes, blood transfusion, and surgical instruments [9]. Furthermore, comparison of 18S rDNA sequences classified as T. equi from multiple locations worldwide indicates greater genetic diversity than expected within a single species. This suggests that multiple taxa are likely to be involved, including additional, so far undescribed and novel species, as was the case for T. haneyi [58].

Animals and Areas of Study
This study was approved by the Universidad Austral de Chile (UACh) bioethics committee under UACh 238/2015. To accurately determine the occurrence of B. caballi and T. equi at the Concepción, Santiago, and Valparaíso racecourses, the required sample size was estimated according to Thrusfield (2007) [70]. The estimated prevalence was 10%, based on a preliminary study in Chile carried out by Vargas et al. (2004), with a precision of 5% and a 95% confidence interval. The minimal sample required was 139 horses from each racecourse. Each owner signed a consent form before the horses' blood samples were taken.
Over 12 months (February 2015 to February 2016), 457 thoroughbred horses had their blood sampled by a veterinary team. One hundred and fifty-one horses came from the "Club Hípico de Santiago" racecourse (Santiago City, Central Chile: 33°27′00″ S, 70°40′00″ W), 139 horses from the "Valparaíso Sporting Club" (Viña del Mar City, Central Chile: 33°0′0″ S, 71°31′0″ W), and 167 from "Club Hípico de Concepción" racecourse (Concepción City, South-Central Chile: 36° 50′00″ S, 73°03′00″ W) ( Figure 5). Blood samples (10 mL) were aseptically taken from all 457 horses via jugular venipuncture, regardless of age, gender, health, and reproductive status, and were placed into Blood samples (10 mL) were aseptically taken from all 457 horses via jugular venipuncture, regardless of age, gender, health, and reproductive status, and were placed into EDTA tubes for molecular analyses and serum tubes for Competitive Enzyme-Linked Immunosorbent Assay (ELISA). Samples were kept in a refrigerator box (4 • C) during transport to the UACh Veterinary Clinical Pathology Laboratory. The serum was separated, and serum and whole blood EDTA were stored at -20 • C until serological and molecular analyses were performed.

ELISA for Babesia caballi and Theileria equi
A total of 150 and 136 thoroughbred horses from Santiago and Valparaíso racecourses, respectively, had their serum samples submitted for ELISA. Horses from the Concepción racecourse were not tested using ELISA as serum samples were not available. Detection of antibodies against B. caballi and T. equi by ELISA was performed using recombinant proteins RAP-1 and EMA-1, respectively, according to the manufacturer's instructions [28][29][30].
OD values were read at 405 nm using an ELISA reader (Dynex Technologies, Chantilly, VA, USA). Serum samples of previously positive and negative horses were used as controls for the ELISA assays [71]. For the sera tested by ELISA, the immunological activity of each serum, the cutoff values, and the ELISA levels (EL) were calculated as described by Machado et al. (2012) [71]. The immunological activity of each serum was calculated by determining the sample to positive serum ratio (S/P), considering positive and negative sera as a reference, using the following equation: S/P = (mean sample absorbance − mean absorbance of negative serum reference)/(mean absorbance of positive reference serum − mean absorbance of negative serum reference). Cutoff values (0.334 and 0.327 for T. equi and B. caballi, respectively) were 2.5 times the mean absorbance value of the negative controls. The readings above the cutoff value were considered positive, according to Machado et al. [72].

DNA Extraction/Purification
DNA extraction and purification from 200 µL of each horse's blood sample were performed using an "E.Z.N.A. Tissue DNA Kit" (E.Z.N.A. Omega BioTek ® , Norcross, GA, USA), according to the manufacturer's instructions, to obtain 50 µL of purified DNA. DNA concentration and purity were determined (NanoDrop ND-1000 spectrophotometer; Thermo Scientific©, Waltham, MA, USA). The absorbance ratio (OD260 nm/OD280 nm) provided an estimate of sample purity, accepting a ratio of 1.8 ± 0.2 as "pure". DNA was stored at −20 • C before performing the PCR assays.

Endogenous Gene
A PCR for the Equus caballus endogenous gene beta-actin protein (ACTB) [73] was performed to check the integrity of the DNA template obtained from blood, as previously described [74] (Table 1).

Piroplasmids PCR DNA Controls
The positive control for B. caballi was generated by gBlock® (Integrated DNA Technolo Theileria equi genomic DNA was obtained from a naturally infected horses from Brazil [29]. Nuclease-free water (Thermo Scientific©, Waltham, MA, USA) was used as a negative control in all PCR reactions.

Nested PCR for Babesia caballi
All 457 samples were tested by the previously described [38] nested (n)PCR for the 48 KDa rhoptry protein gene (RAP-1), specifically for B. caballi. While external BC48F1 and BC48R3 primers (530 bp) were used for the first amplification reaction, internal primers BC48F11 and BC48R31 (430 bp) were used in the second reaction PCR ( Table 3). The reaction mixture was composed of 12.5 µL Gotaq®Green Master Mix (Promega®, Madison, WI, USA), 200 nM of each primer (BC48F1 and BC48R3), and 5 µL of DNA template, brought to a total volume of 25 µL with nuclease-free water (Thermo Scientific©, Waltham, MA, USA). For the preliminary round of amplification, the thermal cycling protocol was as follows: initial denaturation at 94 • C for 5 min, 30 cycles of 94 • C for 30 s, 57 • C for 30 s, and 72 • C for 1 min, with a final extension at 72 • C for 5 min. The same conditions were used for the second round of amplification, except for the set of primers (BC48F11 and BC48R31), using 5 µL of the DNA product from the primary reaction as a DNA template for the nested cPCR for B. caballi. All 457 samples were tested via the previously described nested nPCR protocol [23] to amplify a partial region of the T. equi 18S rRNA gene. A set of external primers (BallF and BallR (814 bp)) and internal primers (BeqF1 and BeqR1 (709 bp)) were used ( Table 3). The reaction mixture was composed of 12.5 µL of Gotaq®Green Master Mix (Promega®, Madison, WI, USA), 300 nM of each primer (BallF and BallR), and 5 µL of DNA template, brought to a total volume of 25 µL with nuclease-free water (Thermo Scientific©, Waltham, MA, USA). For the primary round of amplification, the thermal cycling protocol was as follows: initial denaturation at 94 • C for 5 min, 30 cycles at 94 • C for 30 s, 57 • C for 30 s and 72 • C for 1 min, with a final extension at 72 • C for 5 min. The same conditions were used for the second round of amplification, except the set of primers (BeqF1 and BeqR1), using 5 µL of the DNA product from the primary reaction as a DNA template for the nested cPCR of T. equi.
All the PCR reactions were performed in a T100 TM Thermal Cycler (Bio-Rad, Hercules, CA, USA). The PCR products were separated via electrophoresis on a 2% agarose gel (LE Agarose Seakem®, Lonza, Morristown, NJ, USA) stained with SYBR© Safe DNA gel stain (Thermo Scientific©, Waltham, MA, USA). DNA extraction/purification, cPCR amplification, and electrophoresis were performed in three separate rooms to avoid crosscontamination. The primers used in this study are shown in Table 3.

Molecular Characterization of the Babesia caballi and Theileria equi Agents
Selected positive B. caballi samples presenting strong band intensity in the RAP-1 gene were submitted for sequencing. Additionally, B. caballi and T. equi positive samples with strong band intensity were further tested by targeting the complete 18S rRNA gene (1600 bp), as previously described by Bhoora et al. (2009) [37], for sequencing and molecular characterization (Table 3). Primers NBabesia1F and 18SRev-TB were used in a primary PCR to amplify a fragment of~1600 bp. The reaction mixture was composed of 12.5 µL Gotaq®Green Master Mix (Promega®, Madison, WI, USA), 200 nM of each primer, and 5 µL of template DNA, brought to a total volume of 25 µL with nuclease-free water (Thermo Scientific©, Waltham, MA, USA). The thermal cycling protocol was as follows: 94 • C for 2 min, followed by 40 cycles at 94 • C for 30 s, 60 • C for 45 s, and 72 • C for 1 min. Amplification was completed with a final extension of 72 • C for 7 min.
Three nested PCRs were subsequently performed. The primers NBabesia1F and BT18S3R were used to amplify an 800 bp PCR product at the 5 -end of the gene. The amplification program was the same as that employed for the first PCR, except that an annealing step of 58 • C for 1 min was used. The amplification of the 3 -end of the 18S rRNA gene was accomplished using primers BT18S3F and 18SRev-TB, with an annealing step of 55 • C for 1 min. An internal 800 bp product, which overlapped both the 5' and 3' fragments by 400 bp, was amplified using primers BT18S2F and BT18S2R and an annealing step of 55 • C for 1 min.

Sequencing, Phylogenetic, and Haplotype Analyses
The PCR products (RAP -1 and 18S rRNA genes) were purified with a commercial Silica Bead DNA Gel Extraction Kit (Thermo Scientific®, Waltham, MA, USA), following the manufacturer's instructions, and sent to Macrogen®(Seoul, Korea) for sequencing by the Sanger method in both directions [75]. The nucleotide sequences obtained were analyzed in Bioedit v. 7.0.5.3 (Carlsbad, CA, USA), and consensus sequences were obtained from both the forward and the reverse oligonucleotides. The BLASTn analysis tool was used to compare sequences in the FASTA format with others previously deposited in the GenBank database [76]. Sequences were aligned using MAFFT software [77].
The alignments saved in FASTA mode were transformed into Nexus and Phylip by the Alignment Transformation Environment site, to perform Bayesian analysis using the software MrBayes 3.2.2 in the XSEDE [78] and the maximum likelihood analysis, using the cluster BlackBox RaxML BlackBox 7.6.3 [79], both via the CIPRES portal [80]. Bayesian analysis was performed with 106 generations, with varying numbers of substitution classes according to the evolutionary model found for each dataset and a burn-in = 25%. The bootstrap was accessed with 1000 replicates, and the evolutionary model was based on the Akaike Information Criterion (AIC) using the jModelTest2 software (version 2.1.6) on XSEDE [81]. Phylogenetic tree editing and rooting (via external group) were performed using the Treegraph software 2.0.56-381 beta [82].
Polymorphism nucleotide analysis of 18S rRNA and RAP-1gene sequences was performed using DnaSP v5 [83]. Haplotype diversity (Hd), number of haplotypes (n), and nucleotide diversity (Pi) were analyzed to explore genetic differentiation among the sampled horses. Subsequently, the sequences used in the phylogeny (from Chile and others previously deposited in GenBank) were aligned by the BioEdit Sequence Alignment Editor software [84]. The alignment was re-opened in the DnaSP version 5.10.1 software to group the sequences according to their location and saved in the NEXUS format. Then, the sequences were submitted to the analysis of the TCS network [39], which was inferred using the Population Analysis with Reticulate Trees (popART) program [85].

Data Analysis
To determine the overall EP agent's occurrence in racecourses from Chile, the ELISApositive and the PCR-positive horses were divided by the total number of animals and multiplied by 100. The observed occurrence was expressed in percentages, and the 95% IC was calculated. The occurrence was also calculated per racecourse (Concepción, Santiago, and Viña del Mar).

Conclusions
This is the first molecular and serological survey and genetic assessment of piroplasmids in Chilean Thoroughbred horses. Both Babesia caballi and Theileria equi circulate in the three leading racecourses from Chile, highlighting the impact of the pathogens on the equine industry. Our phylogenetic and haplotype analysis showed a low diversity of Chilean samples compared to the high genetic variability of piroplasmids in horses globally. The sequences we detected in Chile may have originated from Brazil, Israel, or Cuba. Further studies are needed to identify the significant vectors and routes of transmission of the pathogens in this country.