TEXT

Top Abstract Text References

Borrelia turicatae and the closely related spirochetes Borrelia hermsii and Borrelia parkeri are causative agents of New World tick-borne relapsing fever. In North America, relapsing fever occurs primarily in the mountains and semiarid plains of the western United States and Mexico (2). Tick-borne relapsing fever is a zoonotic disease transmissable to humans through the bite of infected argasid and Ornithodoros ticks (11). In addition to the importance of B. turicatae and the relapsing fever spirochetes as causative agents of human disease, the study of these pathogens is important for other reasons as well. The causative agents of relapsing fever and Lyme disease share numerous genetic and physiological traits. One of the more striking shared features of the Borrelia species is their genome, which is comprised of variably sized linear and circular DNA species (3). In addition, the relapsing-fever spirochetes also carry homologs of some Lyme disease spirochete genes such as ospC (19), which is thought to be important in the transmission of the spirochetes from ticks to mammals (24). The ospC homolog in B. hermsii, vmp33, has been demonstrated to be a member of the vmp gene family (6). The existence of homologs of known virulence factors in multiple Borrelia species may indicate that these factors play a genus-wide role in Borrelia pathogenesis. In light of the similarities among Borrelia species, it has been suggested that B. turicatae could serve as a model organism for studying the molecular pathogenesis of Borrelia infections (21). One advantage in utilizing B. turicatae for this purpose is that unlike the Lyme disease spirochetes, this organism reaches high population densities in experimental animal models. This would greatly facilitate studies designed to assess the in vivo expression and function of putative virulence factors.

Cloning and characterization of a homolog of the Lyme disease spirochete rep⁺ gene family in the relapsing-fever spirochete B. turicatae. In this study, we cloned and characterized from B. turicatae a homolog of the Lyme disease spirochete rep⁺ gene family, which we have designated repA (all isolates analyzed in this report are described in Table 1). The recombinant clone carrying this gene (pBt2.2) was serendipitously recovered in the course of screening a B. turicatae (isolate 91E135) genomic library for homologs of the Borrelia burgdorferi sensu lato complex upstream homology box gene family (20). The plasmid library was generated by the ligation of XbaI-digested B. turicatae 91E135 genomic DNA into the polylinker region of pGem-3Zf(+) (Promega). Ligation products were transformed into high-efficiency JM109 cells (Promega) and plated onto Luria-Bertani plates containing ampicillin (100 µg ml¹), IPTG (isopropyl--D-thiogalactopyranoside) (0.5 mM), and X-Gal (5-bromo-4-chloro-3-indolyl--D-galactopyranoside) (80 µg ml¹). Colony lifts were performed with Genescreen membranes as described by the manufacturer (DuPont-NEN). Cycle sequencing with the Sequitherm Excel DNA Sequencing Kit (Epicentre Technologies) of recombinant clone pBt2.2 revealed that it carries a 2.2-kb insert containing a 786-bp open reading frame (ORF). The ORF encodes a putative protein of 262 amino acids (Fig. 1) with a predicted molecular mass of 30.2 kDa. Its deduced pI of 4.69 is unusual in that the median deduced pI of B. burgdorferi proteins is 9.7 (12). A striking feature of this ORF is its central region (nucleotides 212 to 583), which harbors a series of direct repeat elements comprising 47.7% of the ORF. The longest repeat is 28 amino acids in length (Fig. 1). Ten smaller imperfect repeat units of 11 amino acids are also present. The tripeptide KID is present in each of these 10 copies and is perfectly conserved except in the last repeat copy, where it is KIE. Five of the KID(E) motifs are preceded by either T or S. This sequence is identical to that of the putative casein kinase 2 phosphorylation motif, S/T-X-X-acidic amino acid (DNE or Q) (22, 26). The repeat motif-carrying region of the protein is predicted to have a positive Jameson-Wolf antigenic index (13), but Chou-Fasman surface exposure probability predictions (9) for this as well as other regions of the protein are low. Interestingly, the carboxy terminus of the protein is predicted to be the most hydrophobic portion of the protein, with a Kyte and Doolittle hydrophobicity index (14) greater than 1.3. This may suggest that this region of the protein is embedded in the membrane. Since RepA lacks any obvious export signals, it is possible that it may associate with the inner leaflet of the inner membrane. Computer analyses with the Motifs program (Genetics Computer Group) did not reveal the presence of other amino acid sequence motifs that might give additional clues as to the cellular function or location of RepA.

View larger version (36K): [in this window] [in a new window]   FIG. 1.   Nucleotide and amino acid sequence of the B. turicatae 91E135 repA gene. The figure lists the sequence determined from the pBt2.2 clone. The binding sites of all probes and primers used in this study are indicated by double underlines. The actual sequences of the OZ1-PE1 and repAR1 primers are the inverse complements of those underlined. The largest of the amino acid repeats (28 amino acids) present in RepA is indicated by a single underline, and the two copies are denoted by i and ii. The tripeptide repeat motif KID(E) is indicated by bold lettering. The transcriptional start site identified in this report is indicated by an arrow, the ribosomal binding site by bold lettering, and the 10 and 35 elements by underlining.

View larger version (70K): [in this window] [in a new window]   FIG. 2.   Amino acid alignment of the repeat motif region of RepA and its homologs. The amino acid sequences of Borrelia genes exhibiting significant identity with RepA were aligned with the Pileup program and then manually adjusted to minimize gaps. The regions of these sequences that carry the repeat motifs are presented. The tripeptide KID(E) repeat motifs are highlighted by bold lettering and the individual tripeptide KID(E) repeats of B. turicatae 91E135 are numbered.

Southern blot analysis of repA: copy number determination. The repA-related genes carried by B. burgdorferi are either multicopy or exist in the form of gene families (23, 29). To determine if repA is multicopy in B. turicatae, Southern hybridization analyses of restricted DNA were performed with probes targeting different regions of repA. Through these analyses, we also sought to determine if repA or repA homologs are carried by other Borrelia species. To facilitate an accurate comparison of the restriction fragment length polymorphism patterns obtained with the various oligonucleotide probes that were used, multiple sets of XbaI-digested DNA isolated as previously described (18) were run side by side on the same 0.8% GTG-agarose gel (in standard Tris-acetate-EDTA [TAE] buffer [pH 8.0]), transferred onto Hybond N membrane (Amersham), UV cross-linked as previously described (18), and cut to generate identical Southern blots. Oligonucleotide probes (probe binding sites and sequences are indicated in Fig. 1) were 5' end labeled by standard methods with polynucleotide kinase and [³²P]ATP (6,000 Ci/mmol; DuPont-NEN). Hybridizations were conducted at 32°C in a Hybaid hybridization oven (Labnet) by using conditions and buffers that have been previously described (20). All relapsing-fever spirochete species tested (B. hermsii, B. parkeri, and B. turicatae) hybridized with both the repAF1 and repAF2 probes (Fig. 3). These probes target 5' of the central repeat motif region. The repAF1 probe hybridized with several restriction fragments in all hybridization-positive isolates. The repAF2 probe hybridized with multiple fragments in B. turicatae and B. parkeri isolates and with a single fragment in B. hermsii Yor-1. The detection of multiple hybridizing fragments suggests that there are several repA-related genes carried by the relapsing-fever spirochetes. The specificity of the hybridization of the oligonucleotide probes is supported by the fact that many (but not all) of the hybridization-positive restriction fragments bound both repA-targeting probes. This observation also indicates conservation of a significant stretch of the 5' ends of the repA-related genes. A probe (repAR1) targeting the 3' end of the repA-related genes was also employed in hybridization analyses. This probe hybridized with multiple fragments in B. parkeri and one fragment in B. turicatae isolates. The detection of a single hybridizing fragment in the B. turicatae isolates with the repA-R1 oligonucleotide suggests that the 3' ends of the repA-related genes may not be as conserved as the 5' ends.

View larger version (63K): [in this window] [in a new window]   FIG. 3.   Southern hybridization analyses of repA in different Borrelia species. (A) Binding sites of the repA-targeting oligonucleotide probes. The central repeat motif region is indicated by the boxed area. (B) Southern hybridization results obtained with XbaI-digested DNA and various repA-targeting, radioactively labeled, oligonucleotide probes (indicated above each section of panel B). Hybridization conditions were as described in the text. Molecular size standards are shown to the left of each autoradiograph.

Identification of the genetic elements carrying repA-related sequences through Southern blot analysis of 2D CHEF-PFGE-fractionated genomic DNA. To identify the specific genetic elements that carry the multiple repA-related sequences, B. turicatae and B. parkeri DNA was fractionated by two-dimensional contour-clamped homogeneous electric field (Bio-Rad)-pulsed field gel electrophoresis (2D CHEF-PFGE) in 1% GTG agarose gels (15). The algorithm and parameters used were as follows: run time, 20 h 16 min; buffer, 0.5× TBE (Tris-borate-EDTA [pH 8.0]); temperature, 14°C; ramping constant, 1.400; initial switch time, 0.47 s; final switch time, 4.48 s; angle, 120°; gradient, 6 V cm¹. After electrophoresis in the first dimension, the gels were rotated 90° and electrophoresed for 3 h in 0.5× TBE at 80 V (constant field). To facilitate transfer, the gels were stained with 1.0 µg of ethidium bromide ml¹ for 30 min, UV irradiated with 60 mJ of energy, destained, and photographed. Transfer onto Hybond-N membrane was accomplished via vacuum blotting with the VacuGene XL Vacuum Blotting System (Pharmacia) with a vacuum of 55 mbar, a 4- to 6-h transfer time, and buffers described by the manufacturer. As a consequence of the reduced mobility of circular plasmids (relative to linear plasmids) during electrophoresis, which is quite evident upon electrophoresis in the second dimension, linear and circular plasmids can be readily distinguished by 2D CHEF-PFGE (1, 5, 7, 10, 15). Upon hybridization of the blots of the 2D CHEF-PFGE-fractionated DNA, the repA probe hybridized with linear plasmids of 50, 35, 26, and 23 kb in both B. turicatae 91E135 and OZ-1 (Fig. 4). A linear plasmid of approximately 52 kb, present in 91E135 but not OZ-1, also bound the probe. In B. parkeri, linear plasmids of approximately 55, 50, 39, and 28 kb were hybridization positive. To verify that circular plasmids were not migrating along the axis where the linear plasmids migrate, a PCR probe targeting a B. burgdorferi circular plasmid-carried gene (ospC) was used in Southern hybridization of B. burgdorferi 2D CHEF-PFGE-fractionated DNA. We probed for a B. burgdorferi gene since circular plasmid carried genes have not been characterized from B. turicatae. The ospC probe hybridized solely with a plasmid migrating behind the axis of migration of the linear plasmids (data not shown), thereby confirming the differential migration of the plasmid conformations. These data demonstrate that repA-related sequences are carried on a series of linear plasmids in B. turicatae and B. parkeri. The presence of repA on linear plasmids is in contrast to the predominantly circular plasmid localization of the multicopy repA homologs (rep⁺ and ORF-E) of the Lyme disease spirochetes (23, 28, 29). For example, in B. burgdorferi 297, six copies of rep⁺ have been localized to a series of comigrating 32-kb circular plasmids (12, 23). In B. burgdorferi B31, copies of ORF-E are carried on a 50-kb linear plasmid and by circular plasmids of 26, 29, and 30.5 kb (28, 29). Hence, while these different bacterial species carry related genes, these repA homologs are carried on plasmids of different conformation.

View larger version (45K): [in this window] [in a new window]   FIG. 4.   Identification of the genomic elements carrying repA by 2D PFGE and Southern hybridization. DNA liberated from B. turicatae and B. parkeri was fractionated by 2D CHEF-PFGE as described in the text. The DNA was visualized by ethidium bromide staining and was photographed (left panel). The DNA was then transferred onto Hybond N membranes and hybridized with a PCR-generated probe as described in the text. The probe targets almost the entire repA gene and was generated with the repAF1 and repAR1 primers and pBT2.2, the repA-carrying recombinant plasmid, as amplification template. Molecular size standards (lambda monocuts) are indicated between the panels. The isolates analyzed and the direction of electrophoresis in each dimension are indicated.

RT-PCR analysis of repA expression during in vitro cultivation. To assess the transcriptional expression of repA in B. turicatae during in vitro cultivation, Northern hybridization analyses with various probes were performed (data not shown). By this approach, transcript was not detected, raising the possibility that repA either is not expressed during in vitro cultivation or is expressed at levels not detectable by Northern hybridization. To determine if low-level transcription was occurring, reverse transcriptase (RT) PCR was performed with an RT-PCR kit as described by the manufacturer (Perkin Elmer). The RNA template (750 ng per reaction) was purified as previously described (17) except that two rounds of RQ1-DNase digestion were performed to ensure that any contaminating DNA was digested. The repAR1 oligonucleotide served as the primer for first-strand synthesis. To synthesize double-stranded DNA, the repAF1 primer was added. The repAF1-repAR1 primer set would be expected to yield an amplicon of 666 bp and consistent with this, an amplicon of this size was obtained from B. turicatae OZ-1 (Fig. 5). An amplicon was not obtained from B. turicatae 91E135 (data not shown). Amplification products were not obtained from a minus RT, negative-control reaction. A second negative control was devised to further demonstrate that we were not amplifying residual DNA. For this control, a forward primer targeting upstream of repA (303 to 284 bp upstream of the repA start codon) was used in conjunction with the repAR1 reverse primer in an RT-PCR reaction. Since one of these primers targets upstream of the transcriptional start site, amplification would occur only if contaminating DNA were present. Amplification was not observed, providing definitive evidence that the RNA preparations were free of contaminating DNA. To confirm that the RT-PCR amplicon from isolate OZ-1 was in fact derived from repA, the amplicon was cloned and partially sequenced. Partial sequence analysis of the amplicon (340 nucleotides) revealed that it contained four base differences relative to the cloned sequence from B. turicatae 91E135 repA. While it remains to be determined if these sequence differences are real or are artifacts introduced during RT-PCR, it can nonethless be definitively concluded that the amplicon was in fact derived from amplification of a rep transcript.

Identification of the promoter element of repA by RT primer extension. To identify the putative promoter element of repA, RT primer extension was conducted with 5'-end-labeled OZ-PE1 primer, 750 ng of isolated RNA (17) as template, and murine leukemia virus RT (Perkin-Elmer) according to the reverse transcription protocol described above. Extension products were treated with RNase (0.5 µg µl¹) (Boehringer Mannheim), extracted with phenol-chloroform-isoamyl alcohol, precipitated with ethanol, washed with 70% ethanol, vacuum dried, and resuspended in 6 µl of water. Three microliters of stop solution (Epicentre Technologies) was added to the resuspended extension products, and 3 µl was loaded onto a 6% polyacrylamide-8 M urea sequencing gel. An extension product was obtained from B. turicatae OZ-1 (Fig. 5) but not from 91E135. This is consistent with the RT-PCR analyses described above, which demonstrated expression of repA in OZ-1 but not in 91E135. From the size of the extension product, the start site could be mapped to an A residue 16 nucleotides upstream from the translational start codon (Fig. 1). Thirty nucleotides upstream from the transcriptional start site is the sequence TTG CTT, which exhibits identity with other identified Borrelia promoters such as those flanking ospC and ospAB (16, 17). Seventeen nucleotides downstream of the promoter is a conserved 10 or TATA box sequence element, TATACT.

View larger version (29K): [in this window] [in a new window]   FIG. 5.   Transcriptional analyses of repA in B. turicatae OZ-1 cultivated in vitro. (A) To determine if expression of repA occurs during cultivation in vitro, RT-PCR was performed as described in the text. In each reaction, 750 ng of isolated, DNase-treated RNA served as template. In one reaction, RT was omitted [()RT] to verify that all traces of DNA were removed by DNase treatment. A second negative control (5'-primer set) was also performed. In this case, a primer set targeting a region upstream of repA was used. (B) To identify the transcriptional start site of repA, primer extension analyses with the OZ-PE1 primer were performed as described in the text. The resulting primer extension products were electrophoresed in a 6% polyacrylamide-8 M urea gel alongside a sequencing ladder generated with the OZ-PE1 primer and the pBt2.2 recombinant plasmid.

Conclusions. In this report, we describe the cloning and characterization of a linear plasmid-carried gene from B. turicatae designated repA, which is a homolog of the Lyme spirochete rep⁺ genes. These rep-related genes are characterized by the presence of a repeated potential casein kinase 2 phosphorylation site. The description "casein kinase" is broadly applied to at least two classes of ubiquitous protein kinases for which the substrates are not casein but rather a variety of enzymes and noncatalytic proteins that are involved in a variety of cellular functions. The majority of the casein kinase 2 target proteins are highly acidic (as is RepA with a pI of 4.9) and many of the phosphorylated proteins are involved in gene expression and protein synthesis (22). The conservation of sequence and spacing of the casein kinase 2 phosphorylation sites in rep homologs suggests that this amino acid motif may be part of an important functional domain that plays a genus-wide functional role in the biology of the Borrelia.

Nucleotide sequence accession number. The GenBank accession number of the 786-kb ORF sequenced in this study is AF062395.