|Home | About | Journals | Submit | Contact Us | Français|
Persistent infection of a mammalian host by Borrelia burgdorferi, the spirochete that causes Lyme disease, requires specific downregulation of an immunogenic outer surface protein, OspC. Although OspC is an essential virulence factor needed by the spirochete to establish infection in the mammal, it represents a potent target for the host acquired immune response, and constitutive expression of OspC results in spirochete clearance. In this study, we demonstrate that a factor encoded on a linear plasmid of B. burgdorferi, lp17, can negatively regulate ospC transcription from the endogenous gene on the circular plasmid cp26 and from an ospC promoter-lacZ fusion on a shuttle vector. Furthermore, we have identified bbd18 as the gene on lp17 that is responsible for this effect. These data identify a novel component of ospC regulation and provide the basis for determining the molecular mechanisms of ospC repression in vivo.
Borrelia burgdorferi, the causal organism of Lyme disease (4, 8, 46), maintains its complex enzootic life cycle in two different environmental niches, Ixodes ticks and mammalian hosts (32). To achieve this, B. burgdorferi senses key changes in its surroundings and undergoes dramatic adaptive changes in gene expression. As part of this adaptive response, spirochetes in the midguts of infected ticks initiate synthesis of an abundant outer surface protein, OspC, when the ticks take in a blood meal (38, 44, 45). The incorporation of OspC on its outer membrane prepares B. burgdorferi for transmission to a mammalian host, where it establishes persistent infection. We previously demonstrated that B. burgdorferi mutants lacking functional OspC cannot initiate mammalian infection following transmission by infected ticks (26, 48, 50, 52). However, since OspC represents a potent neutralizing target, ospC expression must be downregulated after initiation of infection to avoid clearance of the spirochete by the host's acquired immune response (33–35, 57).
As an infected nymphal tick attaches to a mammalian host and feeds, spirochetes residing in the tick midgut experience changes in temperature, pH, and nutrients which signal a global adaptive response in gene expression through a novel regulatory cascade involving the response regulator Rrp2 and the alternative sigma factors RpoN (also called σN or σ54) and RpoS (also called σS and σ38) (10, 22, 29, 59). In addition to ospC, a number of other B. burgdorferi genes are induced through this RpoN/RpoS signaling pathway during tick feeding (7, 11, 12, 22, 39, 40, 53). However, unlike ospC, many of these genes continue to be expressed during mammalian infection (16). Hence, subsequent repression of ospC is a gene-specific mechanism required to avoid immune clearance.
Xu and colleagues recently described a palindromic sequence immediately upstream of the ospC promoter that represents a potential operator site to which a repressor could bind (55, 56). Although B. burgdorferi mutants lacking this palindromic sequence can initiate mammalian infection, ospC expression is not downregulated in these mutants, and thus they are subsequently recognized and cleared by neutralizing antibodies of the acquired immune response (55). However, the invoked repressor that binds to the operator site and downregulates ospC expression in vivo has not been identified.
In an earlier study, Sadziene and colleagues described several highly attenuated B. burgdorferi clones that constitutively synthesize OspC during in vitro growth, in contrast to the parental B31 strain from which these clones were derived (41). These B. burgdorferi clones had lost many or all linear plasmids during extended in vitro passage but retained the 26-kb circular plasmid 26 (cp26), which carries the ospC gene (41). Synthesis of OspC correlated with loss of a particular linear plasmid, lp17, and ospC expression was highest in clones lacking both lp17 and another plasmid, lp54. It was proposed that lp17 encodes a putative repressor that typically silences ospC in strain B31 during in vitro growth; it was also suggested that further evidence for this repressor could be provided by restoration of lp17 to clones from which it had been lost (41). Although this was technically infeasible when it was proposed in 1993, the genetic system of B. burgdorferi has developed to a stage where displacement and restoration of individual plasmids are now possible (17, 24, 25). In this report, we describe a series of experiments in which we investigated the ability of both full-length and truncated forms of lp17, as well as lp17 gene sequences introduced on a shuttle vector, to repress ospC gene expression.
All B. burgdorferi strains were inoculated from frozen stocks into liquid Barbour-Stoenner-Kelly (BSK II) medium supplemented with 6% rabbit serum (PelFreez Biologicals, Rogers, AR) and were grown at 35°C under 2.5% CO2 (1). All B. burgdorferi strains, Escherichia coli strains, and plasmids used in this study are described in Table 1.
B. burgdorferi was transformed by electroporation as previously described (18, 42). Briefly, 10 to 15 μg of plasmid DNA or genomic DNA was introduced by electroporation into competent B. burgdorferi cells, freshly prepared from an exponential-phase culture. Following electroporation, the cells were resuspended in 5 ml of BSK II and allowed to recover for 18 to 24 h at 35°C. The spirochetes were then diluted in 20 ml of BSK II supplemented with kanamycin (200 μg/ml), distributed at 200 μl/well in 96-well flat-bottom Costar plates (Corning, Lowell, MA), and incubated at 35°C with 2.5% CO2. Alternatively, recovered cells were plated in solid BSK II medium with kanamycin (200 μg/ml) and/or gentamicin (40 μg/ml) and incubated at 35°C with 2.5% CO2.
After 7 to 8 days, 20 μl of bacterial culture from positive wells (identified by a phenol red indicator color change and confirmed by dark-field microscopy) was inoculated into 5 ml of BSK II supplemented with kanamycin and then incubated at 35°C. Typically, approximately 10 of 96 wells were positive for growth, and hence transformants could be considered clonal (60). Total genomic DNA was isolated from outgrowth cultures by use of a Wizard genomic DNA purification kit (Promega, Madison, WI). Transformants were distinguished from spontaneous resistance mutants by PCR with total genomic DNA, using appropriate primers (Table 2). Alternatively, colonies in solid medium were screened directly by PCR to confirm the stable presence of introduced DNA.
Genomic DNA isolated from strain B31-A lp17::kan (14) and plasmid DNAs isolated from four different B. burgdorferi clones (GCB409, GCB413, GCB426, and GCB473) harboring truncated forms of lp17 (3) were used to generate transformants carrying either full-length lp17 or the respective lp17 deletion variant (Table 1). B312 clones carrying full-length lp17 were confirmed by PCR using primers specific for kanamycin and several lp17 genes (Table 2). The deletions in lp17 from the donor GCB strains were confirmed by PCR and Southern blot analysis (data not shown). The deleted forms of lp17 were designated by the respective GCB strains from which they were derived (e.g., pGCB409 is the lp17 variant carried by GCB409). B312 transformants were initially screened by PCR for the kanamycin resistance cassette present on lp17 and then checked by PCR for additional lp17 sequences to confirm introduction of the desired lp17 deletion variant from GCB clones. The complete plasmid content of each B312 transformant was also determined by PCR (19) to confirm that no plasmids other than lp17 had been introduced.
A 2,721-bp fragment of lp17 extending from the 3′ end of bbd14 (nucleotide [nt] 9360) to the 5′ end of bbd19 (nt 12081) was amplified from B31-A3 genomic DNA (primers 11 and 12) (Table 2) and cloned into the pCR2.1 TOPO vector (Invitrogen, Carlsbad, CA). Clones harboring the desired lp17 fragment were identified by PCR amplification and restriction digestion and confirmed by sequencing. The ~2.7-kb insert fragment was excised with XbaI (New England BioLabs, Beverly, MA) and ligated into XbaI-digested pBSV2* vector (5), yielding pBSV2*-7′ (Table 1).
A 767-bp fragment of lp17 (nt 11648 to 10881) comprising the BBD18 coding sequence and downstream sequences (no promoter) was amplified from B31-A3 genomic DNA (primers 18 and 19) (Table 2) and cloned into the pCR2.1 TOPO vector (Invitrogen, Carlsbad, CA). Clones harboring the desired fragment were identified by restriction digestion and confirmed by sequencing. The insert fragment was excised with EcoRI (New England BioLabs, Beverly, MA) and ligated into EcoRI-digested pBSV2* (5), yielding pBSV2*-bbd18 (Table 1). The same bbd18 fragment was also cloned into the shuttle vector under the control of the constitutive flaB promoter, yielding pBSV2*-flaBp-bbd18. Finally, a 1,176-bp fragment of lp17 encompassing the bbd18 gene with 5′- and 3′-flanking sequences (nt 12056 to 10881) was amplified from B31-A3 genomic DNA (primers 17 and 19) (Table 2) and cloned into pBSV2* in a similar fashion, yielding pBSV2*-NP-bbd18 (native promoter) (Table 1).
Plasmid DNA isolated from E. coli harboring pBSV2*, pBSV2*-7′, pBSV2*-bbd18 (promoterless), pBSV2*-NP_bbd18 (native promoter), or pBSV2*-flaBp-bbd18 (constitutive flaB promoter) (Table 1) was used to transform B312. Transformants were selected in liquid BSK II medium in the presence of kanamycin (200 μg/ml) and screened by PCR with specific primers for the kanamycin cassette (6) and bbd18 (Table 2). Shuttle vectors from B. burgdorferi transformants were rescued in E. coli and characterized by restriction digestion to confirm the inserts.
The ospC promoter was amplified from B31-A3 genomic DNA (primers 15 and 16) (Table 2), cloned into pCR2.1 (Invitrogen), and sequenced to confirm the insert. The ospC promoter fragment was excised with BamHI and BspHI and ligated into appropriately digested pBH_lacZBb* (a derivative of pBH_lacZBb with a BspHI restriction site immediately upstream of the B. burgdorferi lacZ reporter [lacZBb]) (27), creating pBHospCp-lacZBb* (Table 1).
Plasmid DNA from E. coli harboring pBHospCp-lacZBb* (Table 1) was used to transform B312 and B312_lp17.8 (Table 1). Transformants were selected in solid medium containing gentamicin (40 μg/ml), and individual colonies were screened by PCR for lacZBb and retention of the kanamycin cassette in B312_lp17.8 (Table 2).
B312 derivatives harboring pBHospCp-lacZBb*, with or without lp17 marked with a kanamycin cassette (Table 1), were grown in BSK II solid medium lacking phenol red and supplemented with the appropriate antibiotic(s). After colony formation, approximately 0.5 ml X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside [20 mg/ml in dimethylformamide]) was added to detect β-galactosidase activity, as described previously (27). Colonies were screened for maintenance of the kanamycin cassette and lacZBb with the appropriate primers (Table 2).
The flaB promoter-bbd18 insert fragment was cut out of pBSV2*flaBp-bbd18 (Table 1) with NotI and ligated into appropriately digested pOK12 (54), creating pOKflaBp-bbd18 (Table 1). E. coli (Top10) was transformed with pOKflaBp-bbd18, pBH_lacZBb*, pBHospCp-lacZBb*, or both pOKflaBp-bbd18 and one of the lacZBb* plasmids. Resulting colonies were screened for the presence of flaBp::bbd18 and/or lacZBb, as appropriate, and streaked onto LB agar plates containing the appropriate antibiotic(s) and X-Gal.
Total RNA was extracted from exponential-phase B. burgdorferi cultures by use of a Nucleospin RNA II kit (Fisher Scientific, Pittsburgh, PA) according to the manufacturer's specifications and was treated with RNase-free DNase I. Synthesis of cDNA was carried out using random hexamer primers and a high-capacity cDNA reverse transcriptase kit (Applied Biosystems, Branchburg, NJ). These reactions were also carried out in the absence of reverse transcriptase to serve as a control for residual DNA contamination. Newly synthesized cDNA was treated with RNase H (Ambion, Applied Biosystems) for 1 h at 37°C to remove RNA-DNA hybrids, and samples were then cleaned and concentrated using a MinElute PCR purification kit (Qiagen, Valencia, CA). Concentrated cDNA samples were quantified by measuring the absorbance at 260 nm and were diluted to 50 ng/μl in DNase- and RNase-free water (Applied Biosystems). Quantitative PCR was performed with 100 ng cDNA, using TaqMan Universal PCR master mix (Applied Biosystems) and primer-probe combinations (Table 2) for the B. burgdorferi flaB (flagellin) (31) and ospC (47) genes on an Applied Biosystems 7900HT instrument. The relative copy numbers of flaB and ospC transcripts were interpolated using a standard curve for each gene target that was generated with purified genomic DNA from 105, 104, 103, 102, and 101 spirochetes. Samples were analyzed in triplicate, and gene expression is reported as the number of ospC transcripts per flaB mRNA copy. The amplification of samples without reverse transcriptase was similar to that for the no-template control. Data sets were compared using GraphPad Prism, version 4.0, for Windows (GraphPad Software, San Diego, CA).
B. burgdorferi protein lysate preparation and separation by 12.5% SDS-PAGE were performed as previously described (51). Gels were run in duplicate and either stained with Coomassie brilliant blue (Sigma, St. Louis, MO) or blotted onto nitrocellulose membranes (Bio-Rad, Hercules, CA). Membranes were blocked overnight with 5% nonfat milk (BD Diagnostics, Franklin Lakes, NJ) in TBS-Tween 20 (20 mM Tris, pH 7.5, 150 mM NaCl, 0.05% Tween 20) and probed with rabbit anti-OspC polyclonal antiserum (1:1,000 dilution) (50), anti-FlaB mouse monoclonal antibody H9724 (1:25 dilution) (2), and rabbit anti-BBD18 antiserum (1:500 dilution) (described below). Next, membranes were incubated with peroxidase-conjugated anti-rabbit or anti-mouse serum (Sigma). Finally, peroxidase activity was detected using Super Signal reagents (Thermo Scientific, Rockford, IL) and X-ray film (LabScientific Inc., Livingston, NJ).
A synthetic 15-amino-acid peptide (CRHFDEQNKTNFNES) matching residues 120 to 133 of the annotated BBD18 protein, with the addition of a cysteine residue at the N terminus for conjugation, was used to generate BBD18-specific antisera in rabbits (Genscript, Piscataway, NJ). Affinity-purified antibodies (Genscript) were used in immunoblot analyses (1:500 dilution) to detect BBD18.
An approximately 1-kb fragment extending through the guaA-ospC intergenic region was amplified from both B31-A3 and B312 total genomic DNAs (primers 7 and 8) (Table 2) and cloned into the pCR2.1 TOPO vector (Invitrogen) in E. coli Top 10 (Invitrogen). Inserts were sequenced using an ABI BigDye Terminator cycle sequencing ready reaction kit with an ABI 3700 DNA sequencer (Applied Biosystems), and the sequences of the guaA-ospC intergenic regions from B31-A3 and B312 were compared using DNA-Star software (DNAStar, Inc., Madison, WI).
Specific downregulation of immunogenic OspC during the early phase of mammalian infection is a key strategy adopted by B. burgdorferi for persistence in the mammalian host (33, 49, 52, 57). Sadziene et al. previously observed that the highly attenuated B. burgdorferi clone B312 expresses an otherwise cryptic ospC gene when lp17 is lost (41), prompting the hypothesis that an lp17-encoded gene product might be responsible for the specific and timely repression of ospC expression in vivo.
In order to address the influence of lp17 on ospC expression, we first determined the plasmid content of strain B312 by PCR, using specific primer pairs for each of the 21 plasmids of strain B31 (data not shown) (13, 19, 23). This analysis confirmed that B312 lacks all B31 plasmids except for cp26, lp54, and several cp32 plasmids (41) (Table 1). We also analyzed the protein content of B312 and confirmed that OspC was synthesized (Fig. 1A, lane 2), in contrast to the case in wild-type (wt) B31, which does not synthesize OspC under typical in vitro growth conditions (Fig. 1A, lane 1) (41, 43). Finally, we analyzed the sequence of the ospC promoter and upstream flanking region (187 bp) of clone B312 and found that it was identical to the wt B31 sequence (data not shown) (23, 36). Thus, although clone B312 constitutively synthesizes OspC during in vitro growth, there are no changes in the 5′-flanking region of the ospC gene that might account for this phenotypic switch.
To investigate the hypothesis that lp17 negatively regulates ospC expression, the entire lp17 plasmid was introduced into B312 by transformation with genomic DNA from a B. burgdorferi strain in which a selectable marker, the kanamycin resistance cassette (6), had been inserted near the left telomere of lp17 (lp17::pKK81) (14). The complete plasmid content of the resulting transformant, B312_lp17.8, was determined by PCR (19) to confirm that no plasmids other than lp17 had been introduced (data not shown). As predicted, B312_lp17.8 no longer synthesized OspC (Fig. 1A, lane 3), consistent with a role for one or more lp17-carried genes in negative regulation of ospC. We also confirmed the presence of OspC on B312 spirochetes by immunofluorescence assay, as well as its subsequent absence when lp17 was restored (data not shown).
The influence of lp17 on ospC expression was analyzed by qRT-PCR with RNAs from B312 clones with and without lp17; the number of ospC transcripts was normalized to the number of flaB mRNA copies for each strain. This analysis demonstrated approximately 10-fold fewer ospC transcripts in the B312 derivative harboring lp17 than in the original B312 clone lacking lp17 (Fig. 1B). These data suggest that lp17 negatively regulates OspC production at the transcriptional level.
We recently developed a lacZ reporter system for B. burgdorferi and used it to analyze ospC gene expression in various B31 clones, including B312 (27). To confirm that introduction of lp17 uniformly inhibited ospC expression in the B312 background, we monitored ospC promoter activity with a lacZ reporter construct, pBHospCp-lacZBb*, using a blue-white colony screen to assay β-galactosidase activity in a large number of transformants. B312 colonies harboring pBHospCp-lacZBb* turned blue after treatment with X-Gal (Fig. 1C), indicating expression from the ospC promoter and production of an active β-galactosidase enzyme, as reported previously (27). However, introduction of pBHospCp-lacZBb* into B312_lp17.8 resulted in most colonies remaining white after incubation with X-Gal (Fig. 1C). We confirmed the presence of pBHospCp-lacZBb* in these colonies by PCR, suggesting that a factor encoded by lp17 repressed lacZBb expression from the ospC promoter in B312_lp17.8. Although we did detect a few blue colonies for B312_lp17.8 harboring pBHospCp-lacZBb* (Fig. 1C), this was a rare occurrence (~1 in 250 colonies). These blue colonies retained lp17, as confirmed by PCR, and had the same ospC promoter sequence as the lacZ reporter construct present in white colonies (data not shown), potentially indicating a mutation in the putative ospC repressor or variation in some other component of ospC regulation (see Discussion).
Our initial observations supported the hypothesis that sequences carried by lp17 repress ospC transcription. In order to narrow down the putative lp17 gene(s) responsible for this effect, we utilized several lp17 deletion variants previously generated by Beaurepaire and Chaconas (Fig. 2A) (3). These terminally truncated forms of lp17 were constructed by insertion of a synthetic replicated telomere at various points along the plasmid and subsequent cleavage by the endogenous telomere resolvase, with recovery of the lp17 fragment that retained essential replication functions and the selectable marker (3). We anticipated that transformation of B312 with a segment of lp17 harboring the putative ospC repressor gene would result in reduced OspC production, similar to what was found with the full-length plasmid. Conversely, introduction of an lp17 deletion variant that lacked the ospC repressor would not diminish ospC expression in B312. Thus, comparing the OspC phenotypes of B312 transformants carrying different segments of lp17 would highlight the region of lp17 sufficient for repression of ospC and thereby facilitate subsequent identification of the sequences responsible for this effect.
We confirmed the structures of the lp17 variants in four clones, GCB409, GCB413, GCB426, and GCB473, whose deletions extended in from either the right or left telomere of the plasmid (Fig. 2A) (3). Results obtained by PCR amplification with primers targeting different regions of lp17 and by Southern blot hybridization with probes for the left and right halves of lp17 were consistent with the predicted sizes and structures of the deleted forms of these plasmids (data not shown).
Transformation of B312 with these truncated forms of lp17 had various effects on OspC synthesis, depending upon which lp17 sequences were introduced (Fig. 2B). The complete plasmid content of each B312 transformant was also determined by PCR (19) to confirm that no plasmids other than lp17 had been introduced. Taken together, these data are consistent with negative regulation of ospC by an internal segment of lp17. Truncations extending in from the right telomere of lp17 were particularly informative, resulting in the complete absence of OspC in B312 strains harboring bbd1 to bbd23 (pGCB409) or bbd1 to bbd19 (pGCB426) yet abundant OspC production in B312 containing bbd1 to bbd14 (pGCB473) (Fig. 2B). These results suggest that the bbd15 to bbd19 segment of lp17 can repress ospC expression.
To test this hypothesis directly, the implicated ~2.7-kb fragment of lp17 was cloned into the shuttle vector pBSV2* and introduced into B312 in the absence of other lp17 sequences. As predicted, B312 transformants harboring the bbd15 to bbd19 segment of lp17 (pBSV2*-7′) no longer synthesized OspC (Fig. 2C and D). These data indicate that a gene(s) or sequences in this region of lp17 either directly or indirectly lead to repression of ospC. The lp17 sequences introduced on the shuttle vector were further limited to those encompassing only bbd18, the largest open reading frame in this region, with or without a promoter (Fig. 2C). As shown by immunoblot analysis (Fig. 2D), expression of bbd18 from either its native promoter or the constitutive flaB promoter inversely correlated with ospC expression in clone B312. Introduction of the promoterless bbd18 gene did not diminish ospC expression. These data demonstrate that expression of bbd18 in clone B312 results in repression of ospC.
We have demonstrated that the ospC promoter-lacZBb reporter accurately reflects expression of the native ospC gene in B312 (Fig. 1C) (27). To investigate whether repression of ospC by BBD18 is direct or mediated through another B. burgdorferi factor, we utilized the ospC promoter-lacZBb reporter in E. coli and introduced constitutively expressed flaBp::bbd18 on a compatible plasmid. Conducting this experiment in a heterologous host in the absence of other Borrelia factors permits a clearer assessment of the mechanism of ospC repression by BBD18. Surprisingly, synthesis of BBD18 did not diminish ospC promoter activity in E. coli, as monitored by colony color (Fig. 3A), or the LacZ protein level and corresponding beta-galactosidase activity in liquid cultures (data not shown). While preliminary, these data suggest an indirect mechanism of repression of ospC by BBD18 in B. burgdorferi.
In this study, we confirmed the hypothesis that the presence of lp17 negatively influences expression of ospC in a highly attenuated B. burgdorferi clone, B312. By restoring all or part of the lp17 plasmid, we showed that an ~2.7-kb region of lp17 is sufficient to repress ospC expression in clone B312, in which all other linear plasmids but lp54 are missing. Furthermore, by introducing a subset of lp17 sequences from this region on a shuttle vector, we demonstrated that expression of the bbd18 gene alone is sufficient to mediate ospC repression in clone B312 of B. burgdorferi. The protein or RNA product(s) encoded by bbd18 exerts a negative effect on ospC expression and could act directly, by binding to the ospC promoter and blocking transcription, or indirectly, by altering the expression of other genes that regulate ospC expression. If working indirectly, BBD18 could either repress an activator or induce a repressor of ospC expression. All of these possibilities would result in an OspC-negative phenotype of B312 when BBD18 is present, whereas OspC synthesis would continue in its absence.
Initial results obtained with an ospC promoter-lacZBb reporter in a heterologous host supported an indirect role for BBD18 in regulating ospC expression (Fig. 3). E. coli cells carrying both the ospC reporter and a constitutively expressed bbd18 gene formed blue colonies, indicating that the lacZBb gene continued to be expressed from the ospC promoter even in the presence of the BBD18 protein. A caveat to this experiment is the possibility that a B. burgdorferi cofactor or modification not present in E. coli could be required for BBD18 to bind the ospC promoter and block transcription. Alternatively, the amount of BBD18 protein made by E. coli (Fig. 3B) may be insufficient or unavailable for efficient repression of the ospC promoter on a multicopy plasmid. As a working model that accommodates the available data, however, we propose that BBD18 indirectly regulates ospC through induction of a repressor that binds to the ospC promoter and blocks transcription (Fig. 4).
During the in vivo infectious cycle, synthesis of OspC is induced during tick feeding and present at the initiation of mammalian infection (45) but subsequently downregulated to avoid immune clearance, presumably through a repressor (15, 16, 20, 33, 34, 37, 52, 55, 57). During in vitro growth, most B. burgdorferi sensu stricto strains do not synthesize OspC unless culture conditions are altered to simulate tick feeding and the early stage of mammalian infection (45), resulting in induction of the Rrp2/RpoN/RpoS regulatory cascade (29, 58, 59). However, clone B312, which lacks many plasmids, constitutively synthesizes OspC without manipulation of culture conditions yet has an ospC promoter identical to that of wt B31. Two unique features of B312 can be invoked to place the role of BBD18 within the context of our current understanding of ospC regulation in wt B. burgdorferi. First, RpoS should be induced in B312. Second, the putative ospC repressor should be made or activated in B312 when lp17 (bbd18) is present. Typically, transient induction of RpoS by shifting the temperature, pH, or growth phase of cultures is needed to induce ospC expression in infectious B31 clones carrying lp17 (Fig. 4), indicating that in a wt background, RpoS is made only under certain culture conditions and that the repressor is not made (56, 58). However, a constitutively activated RpoN-RpoS pathway has also been observed in some ospAB mutants (28). We suggest that regulation of ospC in wt B. burgdorferi reflects a multilayered network of both conditionally regulated and stochastic elements, as illustrated by the mixed phenotypic response of individual spirochetes in an isogenic B. burgdorferi population exposed to the same environmental stimuli (38). The plasmids lost by B312 presumably encode factors that contribute to other levels of ospC and bbd18 regulation, both global and gene specific. The rpoS-induced, ospC-repressed state of B312 carrying lp17 fortuitously mimics the phenotypic state of wild-type B. burgdorferi during persistent mammalian infection, which presents a unique opportunity to investigate ospC repression in vitro.
The bbd18 locus on lp17 implicated in ospC regulation carries a single open reading frame in the annotated B31 sequence (13, 23). Database searches indicate that bbd18 is well conserved among diverse B. burgdorferi sensu lato strains and that related sequences are present in at least some relapsing-fever Borrelia strains but do not identify any homologs of known or unknown function in other organisms. The predicted BBD18 protein has fairly high percentages of charged (29%) and aromatic (11%) amino acid residues, comprising a basic protein with a molecular mass of 25,729 Da and an isoelectric point (pI) of 9.39. Structural predictions suggest a predominantly alpha-helical protein with a few beta-sheets connected by a number of coiled regions. Although the general characteristics and predicted structure of BBD18 are compatible with DNA interaction, analyses of the BBD18 sequence by using the Pfam and InterPro databases (21, 30) do not reveal homology with known DNA binding proteins, DNA binding domains, or transcriptional regulators. Ongoing studies will determine if the entire BBD18 open reading frame is needed for ospC repression in B. burgdorferi and whether BBD18 can bind DNA in a sequence-specific manner.
In this study, we utilized a recently developed lacZ reporter system for B. burgdorferi (27) to demonstrate that bbd18 regulates ospC at the transcriptional level, perhaps indirectly. We are currently using this tool in conjunction with other approaches to delineate which promoter sequences are responsible for BBD18-mediated repression of ospC in B312. Subsequent experiments will determine whether deletion of bbd18 in wt B31 clones results in continued ospC expression in vivo and abrogation of persistent B. burgdorferi infection, as predicted. Ultimately, this combined in vitro and in vivo approach should elucidate the molecular mechanisms by which the essential virulence factor OspC is regulated in the mammalian host.
We thank George Chaconas, University of Calgary, for providing B. burgdorferi strains containing truncated forms of lp17 and Fang Ting Liang, Louisiana State University, for helpful discussions and for sharing unpublished data. We are grateful to Gary Hettrick, Anita Mora, and Austin Athman, Rocky Mountain Laboratories, for assistance with graphics. We thank Frank Gherardini, Paul Policastro, and Tom G. Schwan, Rocky Mountain Laboratories, and all members of the Rosa laboratory for helpful discussions and comments on the manuscript.
This research was supported by the Intramural Research Program of the NIAID, NIH.
Published ahead of print on 22 July 2011.