Morphological, molecular and phylogenetic characterization of Borrelia theileri in Rhipicephalus microplus

This study aimed to perform a morphological, molecular and phylogenetic characterization of Borrelia theileri obtained from infected Rhipicephalus microplus in Brazil. Fifty engorged R. microplus females from cattle in the municipality of Seropédica, Rio de Janeiro, were analyzed for spirochetes by hemolymph smear. Macerated eggs and positive ticks, as well as blood from the bovine infested by these ticks, were analyzed the glpQ, flaB and hpt genes by PCR. The PCR products were purified and sequenced for analysis and construction of a phylogenetic tree. Only 2% (1/50) of the ticks generated a positive result by both smear and PCR. The spiral forms (n = 50) had (media ± SD) a mean length of 19.17 ± 4.12 μm, diameter of 0.2935 ± 0.0469 and number of turns 8.44 ± 2.59. Sequence alignments of the three evaluated genes exhibited 98% similarity to B. theileri isolates, occurring in a clade highly related to B. theileri strain KAT. Egg maceration samples were positive for the three evaluated genes, whereas bovine blood was negative by PCR. This is the most detailed characterization of B. theileri in the Americas to-date, presenting morphological, molecular and phylogenetic data, including the transovarial transmission of the spirochete in the host tick.


Introduction
Species in the phylum Spirochaetes (order: Spirochaetales) are thin, spiral-shaped or wave-like, highly motile bacteria that are visualized by darkfield microscopy.
The subclinical aspects of B. theileri infection were described by Rogers et al. (1999) and Guedes et al. (2008). This species of spirochete is one of the least-described pathogenic tick-borne Borreliae. Historically, this spirochete has been identified in Africa, Australia, and North and South America (MCCOY et al., 2014). In southeast Brazil, Yparraguirre et al. (2007) detected a Rhipicephalus microplus tick infected with Borrelia sp. strain BR, which was highly correlated with B. theileri and B. lonestari. Therefore, this study aimed to characterize the morphological, molecular and phylogenetic features of a B. theileri isolate obtained from naturally infected R. microplus in Brazil.

Ticks and hemolymph smears and examination
Rhipicephalus microplus ticks of the Porto Alegre strain free of Babesia spp. and Anaplasma spp. kindly provided by Dr. Itabajara da Silva Vaz Junior of the Universidade Federal do Rio Grande do Sul in December, 2014(RECK et al., 2009. The ticks were maintained for two generations (January and March, 2015) on infested apparently healthy bovines (Bos taurus), Babesia spp. and Anaplasma spp. free. The animals were acquired from the dairy cattle sector of the Universidade Federal Rural do Rio de Janeiro (UFRRJ) in the municipality of Seropédica, Rio de Janeiro.
A hemolymph screen was performed on 50 engorged females collected after completing their second generation on the infested animal in UFRRJ. The specimens were placed in Petri dishes and fixed in plaster with the ventral region facing upwards. During the non-parasitic phase, the ticks were kept in an incubator at 27 ± 1 °C and a relative humidity above 80%.
On the 8th day of oviposition, hemolymph smears were generated by sectioning the distal region of the first legs of each tick. Hemolymph was dispersed on a 12-cell screen-printed slide for individual evaluation. The hemolymph smears were fixed in methanol and stained with 10% Giemsa and examined under an optical microscope using a 100x oil-immersion objective.

Morphometric analysis
The morphometric analysis of the spiral forms was performed by optical microscopy using the oil-immersion objective, with a coupled photo documentation system (D 'Cell software -Olympus  ). The evaluated parameters were a diameter, total length, and number of spirals for each form.

Molecular analysis
Females that were positive for spiral forms in the smear, their eggs, and the infested bovine blood collected 15 days after tick feeding were submitted for molecular analysis. DNA extraction was performed using a commercial Dneasy  tissue kit (Qiagen  ) following the protocol recommended by the manufacturer.
The PCR mixtures contained 1.0 U of GoTaq  DNA Polymerase (Promega  , Brazil), 1x Green GoTaq  Reaction Buffer, 1.5 mM MgCl 2 , 0.2 mM dNTPs, 1.25 pmoles of each primer and 2.5 µl of DNA template. Samples were initially heated to 95 °C for 5 min to denature the template and activate the polymerase, followed by 40 cycles of denaturation at 95 °C for 60 seconds, annealing at 55 °C (flaB), 53 °C (hpt) or 57 °C (glpQ) for 60 seconds, and extension at 72 °C for 60 seconds, followed by a final extension at 72 °C for five minutes.
PCR products were analyzed by gel electrophoresis (1.5% agarose). For each reaction, one positive (DNA of Borrelia anserina strain AL) and two negative controls (water) were included in the reaction.

Sequencing and phylogenetic analysis
PCR products of the expected size were purified using a PuriLink  PCR purification kit (Life Technology  ) and sequenced using a capillary-type Sanger platform in an ABI 3730 DNA Analyzer (Applied Biosystems, Life Technologies  ). The sequences generated were compared to published data using the NCBI Nucleotide BLAST program. Multiple sequence alignments were performed with sequences obtained from this study and sequences from GenBank using MUSCLE, in the SeaView v.4 program (GOUY et al., 2010). The best-fit evolutionary model was determined using jModelTest version 2.0 (DARRIBA et al., 2012). The Borrelia phylogenetic tree was constructed by the maximum likelihood method with MEGA 7 (KUMAR et al., 2016). The support for the tree nodes was measured using a heuristic search with 1000 bootstrap replicates.

Results
Of the 50 hemolymph smears of engorged female R. microplus ticks analyzed, only one was positive, thus indicating a 2% infection rate. Microscopic analysis of Giemsa-stained smears illustrated a typical morphology for Borrelia spp. (Figure 1). Free forms were observed in the hemolymph ( Figure 1A-E) and forms in contact with the surface of the hemocytes ( Figure 1C, E and F). In total, 50 spirochetes were identified across all fields of the hemolymph smears. The spirochete detected is herein referred to as B. theileri strain C5. The average number of spirals, length, and diameter of 43 spirochetes measured in thick hemolymph smears is shown in Table 1. Longer spirochetes (greater than 27 µm) were suspected to be in the process of dividing or attached to another spirochete in tandem.
The engorged female R. microplus ticks hemolymph positive, and their egg sample was positive for Borrelia by PCR. From the flaB, glpQ and hpt PCRs, gene fragments of 593 bp, 548 bp and 389 bp were generated, respectively. No PCR products were amplified from the DNA in blood samples of the infested cattle. The sequences generated from the PCR-amplified genes were analyzed to determine the similarity to other species in the GenBank Nucleotide database. The partial sequence of the flaB gene showed 100% coverage and 99% similarity to that of B. theileri strain KAT (access number KF569936). It also showed 99% similarity and 99% coverage with sequences of Borrelia sp. strain BR (access number EF141022). For the glpQ gene, there was 98% similarity to B. theileri strain KAT (query coverage 99%) and 100% similarity to Borrelia sp. BR (query coverage 57%). The PCR products of the hpt gene aligned with 99% of the query sequence (query coverage 91%) of B. theileri strain KAT and 90% of the query sequence (query coverage of 99%) of B. miyamotoi strain CT13-2396. The novel sequences were deposited in the GenBank with the following accession numbers MG601737 (flaB), MG601738 (glpQ) and MG601739 (hpt).
The phylogenetic relationship analysis based on the flaB, glpQ and hpt gene sequences obtained in this study formed a clade highly related to B. theileri strain KAT, which was well-supported as being monophyletic (Figures 2, 3 and 4). In the phylogenetic tree of the flaB gene, the clade included Borrelia sp. strain BR, followed by the sequence obtained from this study and B. theileri strain KAT (Figure 2).

Discussion
This study presents the most complete description of B. theileri in R. microplus in the Americas to-date and includes molecular data and some morphological data regarding B. theileri in R. microplus ticks, in addition to the detection of transovarial transmission (TOT).
The description of this spirochete in R. microplus ticks in South America was made only by incidental findings in hemolymph smears (MARTINS et al., 1996;SOARES et al., 2000). In addition to these findings, Yparraguirre et al. (2007) reported a molecular isolate of Borrelia sp. BR in the Southeastern region of Brazil. However, the scarcity of molecular data for B. theileri meant that there was insufficient evidence for accurate identification of the species. However, the alignment of its sequences with B. theileri strain KAT and two other sequences of B. theileri from GenBank possibly identify this spirochete as B. theileri species (MCCOY et al., 2014).
The infection rate identified in our study (2%) was relatively higher than reported in other studies (MARTINS et al., 1996;YPARRAGUIRRE et al., 2007) and (MCCOY et al., 2014) Although the infested animal in this study was negative by PCR, it is possible that it was the source of infection for the positive tick, since B. theileri is usually in low parasitemia (CALLOW, 1967) and therefore would have infected only one of the 50 R. microplus females. McCoy et al. (2014) observed that only one of six Rhipicephalus spp. from the same animal was positive for B. theileri by PCR.
In general, clinical signs of infection in cattle and other animals are mild and variable, but during the latent phase of infection transient increases in rectal temperature up to 39.5 °C and occasional mild depression with anorexia and anemia are observed (MCCOY et al., 2014).
In our study, TOT was observed through PCR amplification of B. theileri genes. These results corroborate Smith et al. (1978) who visualized spirochetes on eggs of R. microplus females infected with B. theileri. In addition, Smith et al. (1978) reported that the bacterium was observed in ovaries, central ganglia and hemolymph, with intense multiplication occurring in hemocytes. In contrast, Martins et al. (1996) were unable to visualize B. theileri in macerated eggs from the infected female. Rollend et al. (2013) reported the occurrence of TOT of B. miyamotoi, the relapsing fever group spirochete in Ixodes scapularis. The study evaluated many I. scapularis tick progenies providing strong evidence that B. burgdorferi sensu stricto TOT does not occur in I. scapularis.
The morphometric analysis of B. theileri strain C5 (Table 1) displays variation in the size and number of spirals, similar to that observed in the other studies of B. theileri presented in the table. It is important to emphasize; however, that organisms of this genus reproduce by transverse fission, which requires elongation of the form and can produce variation in form length (VARELA et al., 2004). Schwarzbach et al. (2015) demonstrated that B. burgdorferi pleomorphism is related to the cultivation conditions, which produces forms ranging from typical (elongated and spiraled) to round bodies. The authors also observed that each morphological variant has distinct biochemical markers, which could influence the pathogenesis of Lyme disease. Here, although B. theileri showed a wide variation in the length and number of spirals, there was no pleomorphism such as described by Schwarzbach et al. (2015). This lack of pleomorphism may be due to being isolated from a single tick, or this accentuated pleomorphism may not be a characteristic of B. theileri.
When first isolating B. lonestari in the ISE6 cell line of I. scapularis, Varela et al. (2004) demonstrated an intimate relationship between the bacterium and the ISE6 cell membrane by electron microscopy, suggesting a process of interaction between B. lonestari and this cell line.
In this study, some spirochetes were in contact with the hemocyte surface of the tick ( Figure 1C, E and F), even in great abundance ( Figure 1F), suggesting that there may also be an interaction between B. theileri and the tick cell. Thus, the methodology used by Varela et al. (2004) may be a candidate approach for a future study of this species.
Phylogenetic analyses of flaB, hpt and glpQ revealed that the sequences generated in this study clearly formed a cluster highly related to other B. theileri sequences in GenBank, including Borrelia sp. strain BR. This result definitively confirms the circulation of this species of Borrelia in South America.
Combining the morphometric and phylogenetic data (but without molecular evaluation), we believe that the spirochete parasitizing the hemolymph of R. microplus females in the American continent belongs to the species B. theileri. In Brazil, spirochetes are suspected of causing disease in humans. The R. microplus tick is among the possible ixodid species involved in the transmission of this bacterium (YOSHINARI et al., 2010). However, to-date, B. theileri has only been found in cattle, horses, goats and sheep (CALLOW, 1967) and there is no record of infection in humans. Nonetheless, as presented in the phylogenetic analysis, B. theileri is highly related to B. lonestari and B. miyamotoi two spirochetes of the relapsing fever group involved in human disease (VARELA et al., 2004;ROLLEND et al., 2013). In addition, the pathogenicity of each strain may vary, as we clearly observe a genetic difference between the different isolates, thus justifying the need for studies of the possible involvement of B. theileri in the etiology of the disease in Brazil.