The relapsing fever spirochete Borrelia turicatae persists in the highly oxidative environment of its soft‐bodied tick vector

Abstract The relapsing fever spirochete Borrelia turicatae possesses a complex life cycle in its soft‐bodied tick vector, Ornithodoros turicata. Spirochetes enter the tick midgut during a blood meal, and, during the following weeks, spirochetes disseminate throughout O. turicata. A population persists in the salivary glands allowing for rapid transmission to the mammalian hosts during tick feeding. Little is known about the physiological environment within the salivary glands acini in which B. turicatae persists. In this study, we examined the salivary gland transcriptome of O. turicata ticks and detected the expression of 57 genes involved in oxidant metabolism or antioxidant defences. We confirmed the expression of five of the most highly expressed genes, including glutathione peroxidase (gpx), thioredoxin peroxidase (tpx), manganese superoxide dismutase (sod‐1), copper‐zinc superoxide dismutase (sod‐2), and catalase (cat) by reverse‐transcriptase droplet digital polymerase chain reaction (RT‐ddPCR). We also found distinct differences in the expression of these genes when comparing the salivary glands and midguts of unfed O. turicata ticks. Our results indicate that the salivary glands of unfed O. turicata nymphs are highly oxidative environments where reactive oxygen species (ROS) predominate, whereas midgut tissues comprise a primarily nitrosative environment where nitric oxide synthase is highly expressed. Additionally, B. turicatae was found to be hyperresistant to ROS compared with the Lyme disease spirochete Borrelia burgdorferi, suggesting it is uniquely adapted to the highly oxidative environment of O. turicata salivary gland acini.

Little is known about the environmental stresses B. turicatae encounters during O. turicata colonisation. Presumably, B. turicatae faces shifts in temperature, pH, nutrient availability, osmolarity, and host-derived reactive oxygen species (ROS) and reactive nitrogen species (RNS), which all significantly depend on whether O. turicata has recently fed or is within a period of prolonged starvation. For example, a study on the related soft-bodied tick Ornithodoros moubata, revealed that the midgut pH can rise to as high as 7.6 during the blood meal and be reduced to pH 6.4-6.8 during periods of starvation (Grandjean, 1983). In contrast, the pH of Ornithodoros saliva is unknown. Previously, we showed that Ixodes scapularis, the hard-bodied tick vector of the Lyme disease (LD) spirochete Borrelia burgdorferi, produced RNS in both its salivary glands and midguts (Bourret et al., 2016). We have found ROS are generated in these tissues as well (data not shown). ROS and RNS appear to be substantial environmental challenges for B. burgdorferi during both blood meal acquisition and during starvation in I. scapularis nymphs, as ROS-sensitive and RNS-sensitive B. burgdorferi strains harbouring mutations in antioxidant defence genes show poor survival compared with their wild-type counterparts (Bourret et al., 2016;Eggers et al., 2011;Li et al., 2007). To the best of our knowledge, the production of ROS and RNS in Ornithodoros ticks has not been reported; however, transcriptomic and proteomic studies have identified a variety of oxidant metabolism and antioxidant defence genes expressed in soft-bodied ticks that transmit RF Borrelia Landulfo et al., 2017; In the following study, we used RNAseq to determine the transcriptome of O. turicata salivary glands to gain insights into the pressures B. turicatae encounters during colonisation of its tick vector.
On the basis of these results, we examined the expression of genes involved in tick oxidant metabolism and antioxidant defences, along with the production of ROS and RNS in O. turicata salivary glands and midguts. This indicated O. turicata salivary glands are highly oxidative environments compared with the tick midguts. Accordingly, we found that B. turicatae was highly resistant to killing by H 2 O 2 when compared with the LD spirochete B. burgdorferi, suggesting the pathogen is uniquely adapted to the oxidative environment of the salivary gland. We recently reported B. turicatae is localised within the acini lumen of salivary glands from unfed O. turicata nymphs (Krishnavajhala et al., 2017), which likely contributes to its ability to be rapidly transmitted from its tick vector to a naïve mammalian host.
To further understand the microenvironment of the O. turicata acini lumen and how it may affect B. turicatae gene expression, physiology, and overall virulence, we assessed the salivary gland transcriptome of uninfected O. turicata nymphs by RNAseq. Using the HiSeq2500 sequencer, a total of 1.51 × 10 7 single-end reads were generated averaging 252 BP in length. The reads were assembled into 10,990 contigs that were annotated. Of the 10,984 contigs, 6,755 had greater than 75% coverage to their best match in the National Center for Biotechnology Information (NCBI) nonredundant database, and 8,183 had e-values by blastp of less than 1 × 10 −15 . The families of proteins are shown in Table S1, and this Transcriptome Shotgun Assembly project was deposited at the DDBJ/EMBL/GenBank under the accession GCJJ01000001-GCJJ010065871. Several antioxidant defence genes were among the most highly expressed genes in the RNAseq analyses, including a glutathione peroxidase and a thioredoxin peroxidase (Table 1). Fifty-five additional genes involved in either antioxidant defence or the generation of ROS and RNS were detected.
Expression patterns for a subset of these genes was confirmed by reverse-transcriptase droplet digital polymerase chain reaction (RT-ddPCR) in both uninfected and B. turicatae-infected O. turicata nymphs ( Figure 1a). The genes included glutathione peroxidase (gpx), thioredoxin peroxidase (tpx), an Mn-dependent superoxide dismutase (sod-1), a Cu 2+ /Zn 2+ superoxide dismutase (sod-2), and a catalase (cat). Although the infection status of O. turicata nymphs was not associated with differences in the expression of these selected antioxidant defence genes, we did observe significant differences in the expression of these genes when comparing their expression in salivary glands and midguts (Figure 1a,b). The expression of gpx was approximately 50-fold higher in the salivary glands compared with the midgut, whereas the expression levels of tpx, sod-1, sod-2, and cat were approximately 10-30-fold lower in the salivary glands compared with the midguts. The observed differences in the

| Expression of genes involved in ROS and RNS production
Previous studies examining the hard-bodied tick I. scapularis have implicated dual oxidase (duox) and nitric oxide synthase (nos) as the primary sources of ROS and RNS production in the tick midgut during infection (Bourret et al., 2016;Yang, Smith, Williams, & Pal, 2014). Our RNAseq analyses indicate that two putative duox genes (duox1 and duox2) and a single nos gene are expressed in O. turicata salivary glands ( Table 1). The expression of each of these genes was confirmed FIGURE 1 Antioxidant gene expression in Ornithodoros turicatae salivary glands and midguts. The expression of antioxidant defence genes were compared by RT-ddPCR in the salivary glands (a) and midguts (b) of uninfected and Borrelia turicatae-infected O. turicata nymphs. The number of copies of each gene were normalised to the quantity of RNA (ng) subjected to reverse transcription. Data represent the mean ± SD of three biological replicates and statistical significance was determined using a two-tailed Student's t-test. No statistically significant differences in gene expression were observed comparing uninfected with infected samples. *p < 0.005 for higher expression in salivary glands compared with midguts. **p < 0.005 for lower expression in salivary glands compared with midguts FIGURE 2 Expression of genes encoding reactive oxygen species (ROS) and reactive nitrogen species-producing enzymes (RNS). The expression of genes involved in ROS and RNS production were compared by RT-ddPCR in the salivary glands (a) and midguts (b) of uninfected and Borrelia turicatae-infected Ornithodoros turicata nymphs. The number of copies of each gene were normalised to the quantity of RNA (ng) subjected to reverse transcription. Data represent the mean ± SD of three biological replicates and statistical significance was determined using a two-tailed Student's t-test.*p < 0.05 compared with uninfected controls. **p < 0.05 for lower expression in salivary glands compared with midguts

| Production of ROS and RNS in unfed O. turicata nymphs
Our transcriptional data, coupled with immunofluorescence data, suggest the salivary glands and midguts of unfed O. turicata ticks are highly oxidative and nitrosative environments, which may be substantial hurdles for B. turicatae infection of these tissues.
Therefore, we probed salivary glands and midguts dissected from Of these, only cdr and sodA have been characterised for their roles in detoxifying ROS (Boylan et al., 2006;Esteve-Gassent et al., 2009).
The expression of cdr and napA appear to be partly dependent on the Borrelia oxidative stress regulator (bosR) in B. burgdorferi (Boylan et al., 2006;Boylan, Posey, & Gherardini, 2003). The B. turicatae genome encodes homologues to each of these genes. Therefore, to determine whether the disparate sensitivities of B. burgdorferi and B. turicatae to ROS and RNS were the result of altered expression of these conserved antioxidant defence genes, we compared their expression in strains grown in mBSK medium by RT-qPCR ( Figure 6). The expression of bosR  FlaB was used as a loading control and also showed similar levels of expression in both B. turicatae and B. burgdorferi samples.

| DISCUSSION
RF Borrelia colonise and persist within salivary glands and midguts of Ornithodoros ticks (Dutton, Todd, & Newstead, 1905;Burgdorfer, 1951  In contrast to what we observed in the salivary glands, the expression of gpx was~50-fold lower in the O. turicata midgut (Figure 1).
Surprisingly, the other antioxidant defence genes we examined were all expressed at higher levels in the midgut compared with the salivary glands. Thioredoxin peroxidase (Tpx) plays a major role in limiting nitrosative stress through the reduction of GSNO to GSH using the reducing power of thioredoxin (Benhar, 2018). The expression of tpx to ROS has been the subject of several studies. Recently, NapA has been described as a copper and iron binding protein, which acts as a metallothionein sequestering copper to prevent cytotoxicity (Wang et al., 2012). Moreover, a napA-deficient B. burgdorferi strain was more resistant to killing by H 2 O 2 , suggesting an inverse relationship of napA expression and resistance to ROS (Li et al., 2007;Wang et al., 2012).
Despite the observed differences in napA mRNA levels in B. burgdorferi and B. turicatae, NapA protein levels were comparable in both strains ( Figure 6). Therefore, the differential susceptibilities of B. burgdorferi  Expression of antioxidant defence genes in Borrelia turicatae and Borrelia burgdorferi. The expression of putative antioxidant defence genes was compared by RT-qPCR using RNA harvested from cultures of B. turicatae and B. burgdorferi grown in mBSK under microaerobic conditions to a cell density of~5 × 10 7 cells ml −1 using flagellin (flaB) as a reference gene (a). Data represent the mean ± SD of three biological replicates. Statistical analysis was performed using a two-tailed, students t-test. *P < 0.01 comparing the expression of B. turicatae genes to B. burgdorferi genes. The expression of FlaB, BosR, and NapA from a set of three biological replicates for B. burgdorferi and B. turicatae were compared by immunoblot (b). The relative fold change in FlaB, BosR, and NapA protein levels in B. burgdorferi and B. turicatae samples were determined by densitometry (c). Data represent the result of three to six biological replicates ± SD. Statistical analysis was performed using an unpaired Student's t-test. *p < 0.05 comparing expression of B. burgdorferi to B. turicatae The data presented here support the hypothesis that RF Borrelia, including B. turicatae, are uniquely adapted to persist in the highly oxidative environment of Ornithodoros salivary glands. Although other RF Borrelia species transmitted by soft ticks will likely exhibit levels of resistance to ROS and RNS similar to those described here, it is unclear if this is true of RF spirochetes that are transmitted by hard-bodied ticks (Armstrong et al., 1996;Barbour, Maupin, Teltow, Carter, & Piesman, 1996;Bunikis et al., 2004;Fraenkel, Garpmo, & Berglund, 2002;Fukunaga et al., 1995;Mun, Eisen, Eisen, & Lane, 2006;Scoles, Papero, Beati, & Fish, 2001)for example, Borrelia miyamotoi and Borrelia lonestari are vectored by Ixodes spp. and Amblyomma americanum, respectively. To date, only a single RF Borrelia species related to B. miyamotoi and B. lonestari has been reported to colonise the salivary glands of Amblyomma geomydae (Takano et al., 2012). Our findings provide evidence that the ability of vector-borne pathogens, such as B. turicatae, to colonise the tissues of their arthropod vectors may be due in part to their ability to adapt to and withstand the ROS and RNS produced in various tissues, including the salivary glands and midguts.  (Barbour, 1984;Battisti, Raffel, & Schwan, 2008

| Bioinformatic analyses
Custom bioinformatic analysis were describe elsewhere (Karim, Singh, & Ribeiro, 2011;Ribeiro, Slovak, & Francischetti, 2017) with modifications. Low-quality reads were trimmed from Fastq files (less than 20) and contaminating adapter primer sequences were removed. De novo assembly of reads was performed using Abyss (with k parameters from 21 to 91 in 10-fold increments) and SOAPde novo-Trans (Birol et al., 2009;Xie et al., 2014). The combined FASTA files were further assembled using an iterative BLAST and CAP3 pipeline, as previously described (Karim et al., 2011). Coding sequences were predicted based on the presence of a signal peptide in the longer open reading frame by similarity to proteins in the Refseq invertebrate database from NCBI, and from proteins from Diptera deposited at NCBI's Genbank and from SwissProt. Reads for each library were mapped for the putative coding sequences using blastn with a word size of 25, 1 gap allowed, and 95% identity or higher was required. Up to five matches were allowed if the scores were identical to the largest score.   PCR (RT-qPCR) reactions were prepared for cDNA samples using the Bullseye EvaGreen and TaqProbe qPCR master mixes (MIDSCI) and oligonucleotide primers listed in Table 2. The RT-qPCR reactions were performed using a CFX Connect Real-time PCR Detection System (Bio-Rad) with cycling conditions of 95°C for 10 s, 59°C for 20 s, and 72°C for 30 s followed by melt-curve analysis to determine the efficiency and specificity of the qPCR reactions. The efficiency of each primer set was determined using CFX Manager software (Bio-Rad) by performing qPCR on 10-fold serial dilutions of purified B. burgdorferi or B. turicatae DNA. Primer efficiencies for each primer set ranged from 95% to 100%. The relative expression of selected genes normalised to the flagellin gene (flaB) was determined using the 2 −ΔΔCT method (Livak & Schmittgen, 2001). The Cq values for flaB showed a coefficient of variance of less than 3% for both B. burgdorferi and B. turicatae cDNA samples.

| SDS-PAGE and immunoblot
Whole cell lysates were prepared from 30 ml cultures of B. burgdorferi and B. turicatae after growth to a cell density of 2 × 10 8 cells ml −1 in mBSK. Cultures were centrifuged at 3,200 g for 17 min to pellet spirochetes. Pellets were washed twice with HN Buffer (10 mM HEPES and 10 mM NaCl at pH 8.0), and whole cell lysates were prepared in Lysis Buffer (4% SDS and 0.1 M Tris, pH 8.0). Protein content was normalised using a BCA. Protein content was normalized using a BCA protein assay per the manufacturer's instructions (Thermo Fisher Scientific, Waltham, MA). Samples were run on SDS-PAGE using a Mini-Tetra system (Bio-Rad, Hercules, CA), and protein bands were visualized using the EZstain system in conjunction with the Gel Doc EZ Imager (Bio-Rad). Membranes were blocked in 5% milk in TBST for 2 h and then washed in TBST. Membranes were then incubated with primary antibodies, including α-FlaB (Thermo Fisher Scientific, Waltham, MA) at a dilution of 1:4000, α-BosR (courtesy of Dr. Frank Gherardini, RML) at a dilution of 1:500, and α-NapA (courtesy of Dr. Frank Gherardini, RML) at a dilution of 1:500. Binding of primary antibodies was detected using an HRP-conjugated secondary antibody, and blots were developed using the ECL chemiluminescence substrate (LI-COR, Lincoln, NE). Blot imaging was performed on the ChemiDoc Imaging system (Bio-Rad, Hercules, CA). Immunoblots were quantified and evaluated for relative fold change using the Image J software. A value of "1" was assigned to B. burgdorferi to determine the relative fold change between B. burgdorferi and B. turicatae.

| Statistical analysis
Data are presented as mean ± standard deviation (SD). To determine statistical significance between multiple comparisons, two-way analysis of variance were performed, followed by a Tukey's post-test.
A two-tailed, student's t-test with the Holm-Sidak method was used to determine statistical significance between two groups. Data were considered statistically significant when p < 0.05.